Re: I was hoping we could have a closer acquaintance Carolyn

[Muhammad Suleiman]

Yeah in terms of what dear?
On Jul 7, 2017 9:25 PM, "Carolyn Meijerhoven"  wrote:
>
> Will you reveal me what do you like?
> http://bitly.com/2sWjzMd

pesky and persistent "driverless" Brother MFC-9340CDW

[Jape Person]

A few weeks ago a CUPS upgrade to our Debian testing systems started 
showing a new driver for our Brother MFC-9340CDW in print dialogs and in 
the CUPS printer list and in the system-config-printer utility.


You'd think that was good news, but we've been unable to find any way to 
make the queue for this "driverless" instance of the printer function 
properly.


The only way we can print with this printer is to do what we were doing 
before the new "driverless" instance of the printer showed up. We add a 
printer to the system via system-config-printer or the CUPS Web browser 
dialog and deliberately select the Brother MFC-9320CW Foomatic or 
Brother Script-3 driver. (That's not a typo. I'm deliberately choosing a 
different model.) Both of those PPDs work. I have to provide a 
deliberately altered name for this instance so users can tell it from 
the one that doesn't work.


The particularly annoying thing about this situation is that I cannot 
delete the "driverless" instance of the printer from CUPS / 
system-config-printer. The instant it is deleted, it is automatically 
re-detected and added back to the printer list. But anyone who chooses 
to print to it is going to get a distorted or garbled printout.


I was able to set a policy in the instance so that only root can print 
to it, so a regular user isn't going to waste time and paper. Still, it 
would be nicer if I could turn off the advertisement that the printer 
and the operating system is providing for the "driverless" instance.


I've found nothing in the printer's menu system or in its Web interface 
that would seem to pertain.


Ideas, anyone?

Thanks,
JP

Re: Live recording

[David Christensen]

On 08/03/17 06:19, Rodolfo Medina wrote:

Jeremy Nicoll  writes:


What makes and models are your mics?


I can't tell, because they're old and I don't keep their specs.  One
of them was given to me about 15 years ago, the other I don't even
remember how I have it.


One more question that might explain the ring -- do the microphones have 
switches?



David

Re: Unable to change mouse acceleration and threshold in Stretch

[Zenaan Harkness]

On Thu, Aug 03, 2017 at 10:59:39PM +0500, Илья Валеев wrote:
> Hello!
> 
> I'm unable to change mouse acceleration and threshold in Stretch.
> 
> It does not depend on DE. There are no error appears. For example, in
> KDE settings applies correctly, ~/.kde/share/config/kcminputrc creates
> and contain properly values, but mouse behavior does not changes.
> 'xset m 3 1' also does not change anything.
> 
> It does not depend on:
> - Stretch 9.0 or Stretch 9.1
> - is it clean install or not
> - mouse
> - is it virtual machine with Jessie host or real system

I use xfce, and a trackball (which requires a couple custom
xorg.conf.d settings to work the way I like).

The man page for xset says under "mouse" that xinput "should be used
if you need device-specific settings."

Do you have any custom Xorg config, or are you just configuring via
kde gui?

(Hopefully someone else can be more helpful.)

Re: Live recording

[David Christensen]

On 08/03/17 05:23, Rodolfo Medina wrote:

The mic input on my PC [may be] stereo.  In fact, it is a laptop, nay a
netbook, doesn't have a line in and it is reasonable that its mic
input is also a line in;


What is the make and model of your netbook?


My Dell Inspiron E1505 laptop has a stereo microphone connector.  When I 
run an OS with good sound support, the audio mixer applet includes a 
means (e.g. checkbox) for selecting between microphone or line-level 
signals.



On 08/03/17 11:15, Rodolfo Medina wrote:

Jeremy Nicoll  writes:


Look at: https://i.ytimg.com/vi/x1q9MXvjDlM/maxresdefault.jpg


All four my mics, all my headphones, all my adaptors, are all of type
2 in the picture, where it says `stereo'.



On 08/03/17 06:19, Rodolfo Medina wrote:

Jeremy Nicoll  writes:


What makes and models are your mics?


I can't tell, because they're old and I don't keep their specs.  One
of them was given to me about 15 years ago, the other I don't even
remember how I have it.  But I've got another teo: they are, each of
them, part of a headphone. They also let hear the recorded sound from
both channels. They are:

http://www.trust.com/it/product/11916-primo-chat-headset

and

http://www.trust.com/it/product/20685-mauro-headset


Do the headsets have one or two phone plugs?  What color or colors? 
What size or sizes -- original 1/4 inch, miniature 3.5 mm, or 
sub-miniature 2.5 mm?


https://en.wikipedia.org/wiki/Phone_connector_(audio)#Modern_connectors


What size are the phone plugs on the non-headset microphones?


David

Re: Network config

[Zenaan Harkness]

On Thu, Aug 03, 2017 at 08:49:05PM +0200, Pascal Hambourg wrote:
> Le 03/08/2017 à 15:52, Zenaan Harkness a écrit :
> > On Thu, Aug 03, 2017 at 08:53:27AM -0400, Greg Wooledge wrote:
> > > But the problem is, various Unix DHCP client daemons do *too much*.
> > > All I want them to do is set the IP address, netmask, and gateway.
> > > I *don't* want them to change the system hostname, or the system
> > > resolv.conf (in which I have hand-placed *our* DNS search domain and
> > > *our* DNS resolvers).
> > 
> > Well, making /etc/resolv.conf read-only, owned by root.root
> 
> ... is just useless. resolv.conf is already owned by root, DCHP client 
> daemons run as root and on Linux systems root
> (uid 0) ignores read/write permissions.
> 
> > uninstalling resolvconf should also solve your problem.
> 
> No, it solves nothing. Without resolvonf the DHCP client will write directly 
> in resolv.conf.
> 
> > If you in this last (laptop) scenario need specific addition of your
> > static nameserver, on top of the DHCP nameservers (e.g. one
> > nameserver might resolve some internal names, the others might
> > recursively resolve internet names)
> 
> This does not work.
> The resolver stops as soon as it receives a positive (record exists) or 
> negative (record does not exist) answer. If it
> receives a negative answer from the first name server, it won't query the 
> next name server.
> 
> All name servers declared in resolv.conf must be equal and provide the same 
> answers, or unexpected behaviour will
> happen. Multiple name servers is only for redundancy and load balancing, not 
> to provide different answers.

What about putting your script lines (copying files around etc) into
your /etc/network/interfaces file?

Given the uniqueness of how you seem to want to do your networking,
perhaps that's the best option to make it less abnormal - looks like
it to me.

Good luck,

Re: Where to install Flash Player?

[Patrick Bartek]

On Thu, 3 Aug 2017 14:01:42 -0400 Thomas George 
wrote:

> I downloaded flash-player-ppapi_26.0.0.137-1_amd64.deb into /opt and 
> used dpkg -i to install it.
> 
> Is this the best way to do it? dpkg -s flash-player-ppapi responds ok 
> installed but dpkg -p flash-player-ppapi doesn't recognize it.

There are a plethora of locations where Flash can be installed
and work.  I've done it manually since my Fedora days.  Download the
tar.gz file directly from Adobe and copy, as root, libflashplayer.so
to /usr/lib/mozilla/plugins/  Never fails to work.

B

port série: liste des vitesses possibles

[Gaëtan PERRIER]

Bonjour,

Est-il possible de connaître la liste des vitesses disponibles pour un port
série, qu'il soit physique ou un adaptateur USB/Série ?

Gaëtan


pgp8_PybfMFLm.pgp
Description: PGP signature

Re: Sudden death of bluetooth headphone connectivity

[Henrique de Moraes Holschuh]

On Fri, 04 Aug 2017, Mark Fletcher wrote:
> suddenly stopped working. I didn't do an apt autoremove last Sunday, but 
> may have done one the Sunday before -- which would have been reckless 

The log files are at:
/var/log/apt*
/var/log/apt/*

Failing that:
/var/log/dpkg*

It will list everything that was updated, installed, removed, or purged.

-- 
  Henrique Holschuh

Re: Sudden death of bluetooth headphone connectivity

[deloptes]

Mark Fletcher wrote:

> If anyone can remember which package that is, I can check if it is
> installed... The only thing is I am not sure how it would have got onto
> the system if it isn't depended on and I didn't install it manually. And
> if it IS depended on, or I did install it manually, apt autoremove
> should not have removed it...

to start with

pulseaudio-module-bluetooth - Bluetooth module for PulseAudio sound server

Re: Live recording

[deloptes]

First of all you need to get basic knowledge of signal and audio processing

One good way to understand things 8especially about electricity is water and
pipes.

Now your mic is a one bucket full of water and you have to pipes (left and
right) ... where does the water flow?

Rodolfo Medina wrote:

> With my two microphones, say mic1 and mic2, I did a little experiment: I
> recorded my voice with mic1, using sox, then listened to the so created
> .wav file and the sound was heard only on the left channel.  Instead,
> doing the same with mic2, the sound was heard on both channels.  How
> should we conclude?  It seems to me that we should conclude that:
> 
> 1) the mic input on my PC is stereo.  In fact, it is a laptop, nay a
> netbook, doesn't have a line in and it is reasonable that its mic input is
> also a line in;
> 

there is no stereo mic - keep in mind - one bucket full of water - not two

> 2) mic1 is mono;
> 
> 3) mic2 is stereo.
> 

now way - such thing does not exist - one can not be two ... even easier ...
you have two ears and one mouth ... so you hear stereo, but you speak mono

> Do you agree?  But then, if it is so, my rough home made live piano
> record, done using mic1 and mic2 plugged together in a
> one-male--two-female splitter, is actually stereo...?
> 

it really depends how this splitter is wired. What I post you before is what
you need - a splitter that splits the stereo L/R to a single L and single
R.

Stereo jack has 3 contacts mass and L,R while mono has 2. So I conclude
something is wrong with your wiring. That's it.

get a propper splitter and try the line in

Re: Sudden death of bluetooth headphone connectivity

[Mark Fletcher]

On Thu, Aug 03, 2017 at 10:39:03PM +0200, deloptes wrote:
> Mark Fletcher wrote:
> 
> > 
> I vaguely recall there was a package that needs to be installed for the
> audio to work, but it does not show into dependencies, so it is not pulled
> with bluez or pulse audio (this is based on jessie storie), while it is
> providing the link between them. Perhaps you did autoremove not used
> packages and lost it.
> 
> Unfortunately I upgraded as well and never had the time to see if audio via
> bluetooth works in stretch
> 
This has the ring of common sense to it -- it would explain why it 
suddenly stopped working. I didn't do an apt autoremove last Sunday, but 
may have done one the Sunday before -- which would have been reckless 
even for me just before going on a business trip during which I wanted 
to be able to access my machine remotely [albeit not the bluetooth 
parts], but I suspect I did it anyway... I certainly did one recently, 
just can't remember exactly when, and hence can't be sure whether I have 
used bluetooth headphones successfully since...

If anyone can remember which package that is, I can check if it is 
installed... The only thing is I am not sure how it would have got onto 
the system if it isn't depended on and I didn't install it manually. And 
if it IS depended on, or I did install it manually, apt autoremove 
should not have removed it...

To someone's point, Google'S coming up fairly blank on this one...

Mark

Re: Re: Sudden death of bluetooth headphone connectivity

[Mark Fletcher]

On Thu, Aug 03, 2017 at 08:25:27PM +0100, dekkz...@gmail.com wrote:
> 
> Mark,
> 
> What BT chipset are you using, have you acccidently disabled it in BIOS?
> Nothing shows up in google about protocol not available but there was a bug
> in gnome 3.24 and pulseaudio 10 with gnome/GDM where GDM starts its own
> pulseaudio instance.
> https://wiki.archlinux.org/index.php/Bluetooth_headset#Gnome_with_GDM
> 
> I don't run gnome on debian but i do on arch, i didn't find the gnome BT
> stack particularily stable so i used blueman-applet as a backup in the cases
> gnome BT wouldn't work with my FSL360. see blueman & bluez-utils packages in
> stretch.
> 
> Blueman-applet allows you see if the device is trusted & paired and to
> perform those functions if you wish. It also identifies the BT adapter in
> use which should appear in lsusb.
> 
> If that's no further help look up the Arch wiki section on bluetoothctl
> [from bluez-utils]. 99% of the time the issues i had were BT getting
> confused about adapters.

Some more info I just discovered.

First of all, it's not just a2dp-sink protocol that is reported missing. 
I just tried to connect my iPhone which I set up back in the Jessie days 
as an a2dp SOURCE so I could play music from it through my PC speakers, 
and got the same behaviour and an error in the systemd journal saying 
a2dp-source not available. IE same error but source instead of sink. To 
me, this implies the problem is not specific to a2dp-sink protocol.

Now the variable here is that it was way more of a fiddle than I 
expected it to be to get playing music from my iPhone through the 
computer's speakers to work in the first place, years ago, and I haven't 
tried since the upgrade from Jessie->Stretch, so THIS PARTICULAR task 
could have been broken since the upgrade and I wouldn't have known. It's 
far more important to me to get my headphones working again than it is 
to get the iPhone able to play music through the computer's speakers.

To your point about the bluetooth potentially being disabled, I don't 
think so because: A) the bluetooth hardware on this circa-2009 
self-built Core i7-920-based PC with an ASUS P6T motherboard is a USB 
bluetooth dongle about which the BIOS knows nothing, and B) I tried the 
only bluetooth operation I could think of that didn't involve audio, 
trying to connect to my iPhone as a network provider, and while the 
connection failed it did so at the stage of the computer asking it for 
an IP address, NOT at the stage of trying to connect via bluetooth. In 
fact I can clearly see it report success connecting to the phone in the 
log, followed by a series of error messages relating to DHCP, and then 
the connection gets dropped. In terms of visual effect, in this case the 
slider button stays in the ON position for several seconds before going 
back to OFF, in contrast to the audio-based bluetooth attempts which 
fail instantly.

The phone isn't set up to provide a network, so it is no surprise to me 
that that operation failed, but the point is it DIDN'T fail at the 
bluetooth stage.

I guess the next thing to try is your blueman suggestions. It strikes me 
as odd that it would be working fine and then suddenly break though.

I had always thought that blueman IS the Gnome bluetooth applet -- 
guess that was my misunderstanding.

By the way, I failed to mention in my original post, but in case anyone 
is wondering the audio on my PC is basically working fine, if I watch a 
video or something on the computer using the computer speakers, the 
audio is fine. So pulse, audio config on my PC, permissions etc, all are 
demonstrated OK. No problems until I bring the bluetooth devices into it.

Mark

Re: Flash Player Problem

[MENGUAL Jean-Philippe]

Hi,



Do you absolutely want to use free software for Flash? Otherwise,
pepper-flash package is a good solution, up-to-date and working.


Best regards,



Le 03/08/2017 à 19:38, Thomas George a écrit :
> An attempt to log in to the Bridgebase website fails.
>
> Login using firefox ESR results in an all blue screen. The firefox
> plugin shockwave flash is installed.
>
> My system is debian stretch using the LXDE desktop. I have also
> installed gnash, gnash-common, and browser-plugin-gnash in an attempt
> to solve this problem.
>
> Login using Chromium browser fails with note to install flash player
> BUT dpkg -s flash-player-ppapi reports installed ok.
>
> bridgebase is the only website that has this problem. I have contacted
> their support
> and they suggest problem is changes in flash or in the browsers.
>
> Can you help?
>
>

-- 
signature_jp_2
Logo Hypra  JEAN-PHILIPPE MENGUAL
DIRECTEUR TECHNIQUE ET QUALITÉ
102, rue des poissonniers, 75018, Paris
Tel : +331 84 73 06 61  Mob : +336 76 34 93 37

jpmeng...@hypra.fr 
www.hypra.fr 
Facebook Hypra  Twitter Hypra
 Linkedin Jean-Philippe

Re: Coucou! vas-tu me dire un petit 'salut'? Melanie

[abdelkrim daghar]

bsr Melani comment vatu

> Envoyé: jeudi 3 août 2017 à 17:28
> De: "Melanie Kherchoof" 
> À: debian-user@lists.debian.org
> Objet: Coucou! vas-tu me dire un petit 'salut'? Melanie
>
> 
> 
> Viens qu'on discute un peu toi et moi. 
> http://bit.ly/2u4im6W
>

Re: Lenovo Ideapad 110 algún truco?

[Polo Oyarzún]

Sólo para documentación y por si alguna vez se encontrara una salida
instalé la imagen iso desde este repositorio unofficial escogiendo la que
viene con kde.
Al hacerlo terminó la instalación y al probarlo instalado se cuelga y no
funciona en lo absoluto. Inicialmente podría decir que debian no es
compatible con lenovo ideapad 110. Si alguien ha tenido tiempo de hacer
otras pruebas y puede indicar lo contrario estaré feliz de saberlo.

De momento mantengo el laptop con linux mint 18.2 y funciona de las mil
maravillas.



El 21 de julio de 2017, 21:35, Oswaldo Ortiz 
escribió:

> Puedes intentar usando los siguientes live cds e instaladores ya que
> incluyen el non-free firmware que no trae el instalador normal.
>
> https://cdimage.debian.org/cdimage/unofficial/non-free/
> cd-including-firmware/
>
> --
>
>  Original Message 
> Subject: Lenovo Ideapad 110 algún truco?
> Local Time: July 21, 2017 2:05 PM
> UTC Time: July 21, 2017 8:05 PM
> From: correo...@gmail.com
> To: Lista. Debian 
>
> Muy estimados miembros de la lista.
> Intenté instalar debian en un lenovo ideapad 110 y como live no arranca y
> con el instalador me aparece un mensaje que indica que para que funcione
> necesita una larga lista de paquetes propietarios y que si los tuviera
> inserte el disco.
> Probé con ubuntu y reconoció todo el hardware de inmediato. Lo mismo con
> linux mint.
> Sólo quería saber si de casualidad a alguien le ha pasado lo mismo y ha
> logrado instalar debian en ese modelo de notebook.
> La verdad es que no tengo tiempo de hacer muchas pruebas ni de
> investigaciones muy largas por lo que si alguien tiene una receta sencilla
> y rápida agradecería me la comparta para dejar mi notebook con el debian
> original y no con sus derivados.
>
>
>

Re: Créer un fichier grub.cfg correct

[Pascal Hambourg]

Le 03/08/2017 à 20:37, andre_deb...@numericable.fr a écrit :

Comment recréer un fichier "grub.cfg" correct ?

Sur le disque dur, partition ext4 : sda2.
Deux sauvegardes ext4 sur un 2ème disque dur : sdb1 et sdb2.


Sauvegardes de quoi ? Sous quelle forme ?


Je veux booter sur sdb1 et me retrouve sur sda1.


Donc sdb1 contient un système ? Je croyais qu'elle contenait une sauvegarde.

Re: Hébergeur à conseiller (pro-debian ?)

[daniel huhardeaux]

Le 03/08/2017 à 18:56, kaliderus a écrit :

Bonjour,

Question simple :

Quel hébergeur me conseillez-vous (ou déconseillez) et pourquoi ?

Besoins :
- 1 nom de domaine
- 1 machine sur laquelle j'ai totalement la main, à priori physique,
donc rien de mutualisé
- échanges illimités
C'est un cadre professionnel, donc si on peut payer par chèque ou
virement chaque moi ça m'arrange (pas de carte de paiement :-( )

Merci la liste.
K



Ayant testé OVH et OnLine, je suis passé chez Hetzner en Allemagne. Pas 
photo, paiement par virement accepté avec les contraintes que cela 
engendre (pas chèque, c'est un mode de paiement franco-français). Je 
suis en paiement CB, mode professionnel également.


--
Daniel

Re: Hébergeur à conseiller (pro-debian ?)

[Haricophile]

Le Thu, 3 Aug 2017 20:26:46 +0200,
Dominique Asselineau  a écrit :

> OVH acceptent-ils les chèques ?  C'était dans la demande initiale.
> 
> Dominique

Tous les hébergeurs acceptent les virements, c'était aussi dans la
demande initiale et a mon humble avis c'est quand même un meilleur
outil : 
Rapide, sans envoi postal, ne provoquant pas de frais en cas de
découvert non autorisé (simplement bloqué), renouvelable
automatiquement ou planifiable pour ne pas oublier les échéances,
pouvant être stoppé n'importe quand (contrairement à la carte bleue),
gérable par internet voire interfacé avec les logiciels de gestion... 
Bref, un truc tout a fait adapté aux entreprises, mais aussi aux
particuliers dans ce genre de transaction.


-- 
haricoph...@aranha.fr 

Re: Arial vs. Helvetica.

[Felix Miata]

pe...@easthope.ca composed on 2017-08-02 06:26 (UTC-0700):

> I want to specify a variable pitch font in a wiki. The font should be 
> widely available and acceptable to commonly used browsers.  
> https://en.wikipedia.org/wiki/Arial explains, "It [Arial] was created 
> to be metrically identical to the popular typeface Helvetica, with all 
> character widths identical, so that a document designed in Helvetica 
> could be displayed and printed correctly without having to pay for a 
> Helvetica license."  That suggests that Arial is a good choice.  
> Conversely, font substitution appears to be handled well in many 
> contexts and licensing might no longer be a concern.

> What is the conclusion.  Should I specify, Arial or Helvetica or 
> something else?

Something else:

sans-serif

For the first Internet decade at least, Arial was the default sans-serif font on
Windows, and Helvetica(Neue) was the default sans-serif on Mac. Arial is a
functional clone of Helvetica, both of which are scalable font types. The
defaults may have since changed, but those two will still be available, and if
the defaults were changed, either by Microsoft or Apple, or /users/ (which has
always been a principle of the personalization of personal computing devices),
the change would have been for the better, meaning good if not excellent choice
for the user's environment. Linux users have often been able to install the
Windows fonts, but those who haven't have nevertheless been supplied with metric
equivalents (e.g. Liberation Sans), so also find their defaults to be suitable.
Licensing issues are only of concern on the user end, unless your documents are
using embedded fonts that are limited by license.

Wiki software tends to differ rather little from styles create by stylists
themselves by including similarly long styles in their various themes.

All that said, too few web stylists style fonts so simply. It's typical to see
font declaration lists too long to fit an 80 character window, over-controlling,
which typically means the optimal font for the stylist, and too commonly
blocking the font that the user or his OS vendor has determined to be optimal.
Additionally, all modern browsers provide stylists support for web fonts
downloaded along with the markup, styles and scripts, instead of being limited
to fonts already installed on the users' systems. An example would be

included in the HTML document head and 'russo one' among the document's styles.

It ultimately comes down to how much control you feel is appropriate to impose.
Just remember users can veto everything by disabling styles, or override
particular styles via browser add-ons or CSS customization. In theory, CSS
styles are merely suggestions.
-- 
"The wise are known for their understanding, and pleasant
words are persuasive." Proverbs 16:21 (New Living Translation)

 Team OS/2 ** Reg. Linux User #211409 ** a11y rocks!

Felix Miata  ***  http://fm.no-ip.com/

Re: Sudden death of bluetooth headphone connectivity

[deloptes]

Mark Fletcher wrote:

> 
> [Full disclosure -- I previously had a different problem on upgrading to
> Stretch from Jessie -- see this link for that problem and a link to its
> solution --> https://lists.debian.org/debian-user/2017/06/msg01196.html
> I have never been particularly happy with the solution I linked to in
> that thread, precisely because I was worried it would be damaged by
> updates, but it seems to be intact at this point, so I don't think that
> is what the problem is here]
> 
> Thanks in advance if anyone can help me diagnose the sudden loss of the
> a2dp-sink protocol. The headphones work fine with my iPhone, so the
> problem seems to be with the computer, with which they were working fine
> until a bit over a week ago.

I vaguely recall there was a package that needs to be installed for the
audio to work, but it does not show into dependencies, so it is not pulled
with bluez or pulse audio (this is based on jessie storie), while it is
providing the link between them. Perhaps you did autoremove not used
packages and lost it.

Unfortunately I upgraded as well and never had the time to see if audio via
bluetooth works in stretch

re: Coucou! vas-tu me dire un petit 'salut'? Salome

[Christian ALLION]

 

 

   STOP  STOP

 

 

 

 

> Message du 03/08/17 18:20
> De : "Salome Dvitharana" 
> A : debian-user@lists.debian.org
> Copie à : 
> Objet : Coucou! vas-tu me dire un petit 'salut'? Salome
> 
> 
> 
> Viens qu'on discute un peu toi et moi. 
> http://bitly.com/2vtz3vQ
>

Re: Hébergeur à conseiller (pro-debian ?)

[Cyrille]

> OVH acceptent-ils les chèques ?  C'était dans la demande initiale.
oui
https://www.ovh.com/fr/support/documents_legaux/payment.xml#cheque

Re: Créer un fichier grub.cfg correct

[oblivion.ikiub]

Bonjour, 

As tu verifié ton dossier /boot et le fichier /etc/default/grub s'il n'y a pas 
d'anomalies/coquilles  ?  Il me semble que update-grub génère le fichier 
grub.cfg en ce basant sur /boot


Envoyé de mon Meizu PRO 6

 Message original 
Expéditeur : andre_debian 
Heure : jeu. 8/3 20:37
À : debian 
 Objet : Créer un fichier grub.cfg correct

>Comment recréer un fichier "grub.cfg" correct ?
>
>Sur le disque dur, partition ext4 : sda2.
>Deux sauvegardes ext4 sur un 2ème disque dur : sdb1 et sdb2.
>
># update-grub
>me créé un fichier "grub.cfg" avec plein d'erreurs, contenant trop de 
>lignes inutiles, avec des UUID ne correspondant pas aux partitions,
>Exemple :
>$menuentry_id_option 'gnulinux-simple-a8a150e7-77f0-4e27-bf93-33ab8932e427'
>root=UUID=f9cafbc8-5b78-4b8a-9e5f-f50697ab33f4
>
>Deux UUID différents pour la même partition !!!
>Je veux booter sur sdb1 et me retrouve sur sda1.
>
>Je désire un "grub.cfg" contenant par partition ceci :
>un boot graphique + un boot recovery (single),
>soit environ ~6 lignes,
>avec toutes les lignes UUID qui correspondent bien aux 
>/dev/sda(x) et sdb(x),
>et supprimer ces moults boot à foison.
>
>Y a t-il moyen de purger grub, car j'ai l'impression que
>grub utilise d'anciennes données, soit repartir à zéro.
>
>En supprimant grub.cfg, pareil, fichier pelin d'erreurs.
>
>Merci, André
>

Re: Hébergeur à conseiller (pro-debian ?)

[Gaëtan PERRIER]

Le Thu, 3 Aug 2017 20:14:41 +0200
"Ph. Gras"  a écrit:

> Bonsoir,
> 
> >>> C'est un cadre professionnel, donc si on peut payer par chèque ou
> >>> virement chaque moi ça m'arrange (pas de carte de paiement :-( )
> 
> Gandi, OVH… ou n'importe quel autre si possible d'une capacité capitalistique
> équivalente à la boîte où tu bosses, pour éventuellement faire pression en
> cas de besoin. Si possible pas trop loin, histoire de pouvoir s'y
> déplacer ;-)
> 
> Gandi ne vend que des machines virtuelles onéreuses dès que tu commences
> à avaler du trafic. J'ai fait récemment basculer une assoce vers un Kimsufi,
> car ce dernier revient aussi cher que les 2 disques qu'elle a chez Gandi,
> pour une capacité 3x plus importante.
> 
> OVH est une grosse boîte à présent, avec un service commercial dégueulasse.
> 
> À partir de 15-20 € par mois, tu as maintenant pléthore de fournisseurs…
> 
> Bonne continuation,
> 
> Ph. Gras

Que pensez-vous de https://onetsolutions.net ?

Gaëtan


pgpK5vStte45w.pgp
Description: PGP signature

Re: Live recording

[Rodolfo Medina]

Doug  writes:

> On 08/03/2017 08:19 AM, Rodolfo Medina wrote:
>> Jeremy Nicoll  writes:
>>
>>> On Thu, 3 Aug 2017, at 13:23, Rodolfo Medina wrote:
>>>
 3) mic2 is stereo.
>>> Not necessarily.  It might be a mono signal that's been duplicated across
>>> two outputs.  It's only 2-channel if it can record two separate sounds at
>>> the same time.
>>>
>>> You can get microphones which, in a single unit, record more than one
>>> channel at once.  Their labelling or artwork on them normally makes it
>>> clear that they have L and R (or more than that) inputs.
>>>
>>> What makes and models are your mics?
>>
>> I can't tell, because they're old and I don't keep their specs.  One of them
>> was given to me about 15 years ago, the other I don't even remember how I
>> have it.  But I've got another teo: they are, each of them, part of a
>> headphone.  They also let hear the recorded sound from both channels.  They
>> are:
>>
>>   http://www.trust.com/it/product/11916-primo-chat-headset
>>
>> and
>>
>>   http://www.trust.com/it/product/20685-mauro-headset
>>
>> Rodolfo
>>
>>
> I looked at both of the links. Note that there are two connections for each
> device: headset (earphone) connection and mic connection. The headset
> connection will be
> stereo, and the mic connection will be mono. If there is talkback (i.e., you
> can hear the mic output on the headphones,) the mic sound will appear in both
> earpieces,
> but will not be stereo--it will just be the same sound in both earpieces.
>
> You can't get something for nothing. If you want stereo, you will have to
> have two mics or a mic with two separate microphone elements aimed in two
> (left and right) directions. Such a device will have a plug with three
> connections on it: left, right, and ground. Your PC will not be able to
> handle such a microphone! You would need a stereo preamplifier to plug that
> mic into, and then you would plug the stereo output of the preamp into the
> LINE IN jack of the PC.
>
> I would hope that this information will settle the question!
>
> --doug

why do you get angry...?  there's nothing to get angry, I think...

Re: Re: Sudden death of bluetooth headphone connectivity

[dekkzz78]

On 08/03, Mark Fletcher wrote:

On Thu, Aug 03, 2017 at 02:03:36PM +0100, stuart watt wrote:

On 08/03, Mark Fletcher wrote:
> Hello the list!
>
> Suddenly, earlier this week, my bluetooth headphones stopped working
> with Stretch.
>
> I update weekly, usually on Sundays, and I am not sure because I wasn't
> paying attention whether I had successfully used my headphones before
> the last update I did last Sunday.
>
> They definitely were working properly one week before that, when I used
> them with the computer just before going off on a week-long business
> trip.
>
> Anyway, I use Gnome. When I want to connect the headphones I turn them
> on, click on the network-like icon in the top right of the screen, click
> on "Not in Use" beside the Bluetooth icon, and then click on "Bluetooth
> Settings". (Is there a more efficient, involving fewer clicks, way to
> connect, by the way?)
>
> In Bluetooth Settings I can see my headphones as "BeoPlay H8". If I
> click on them it brings up the window with the slider switch to connect
> them. Normally, what would have happened is that clicking that switch
> would have connected the headphones. WHat happens now is the switch goes
> very briefly to the ON position, and then goes straight back to the OFF
> position. Removing / forgetting the device and re-pairing leads to the
> same result -- pairing is successful, but cannot connect.
>
> With journalctl -f running in a terminal while trying to connect I see
> the following at the moment of clicking the slider switch to connect the
> headphones: a2dp-sink profile connect failed for AA:BB:CC:DD:EE:FF:
> Protocol not available
> 
>
> Now, call me a super-sleuth, but I suspect that Protocol Not Available
> error could be something to do with the problem ;)
>
> To re-iterate, this was working until recently, and all I've done is
> install regular updates using apt update / apt upgrade.
>
> Where should I start looking?
>
> [Full disclosure -- I previously had a different problem on upgrading to
> Stretch from Jessie -- see this link for that problem and a link to its
> solution --> https://lists.debian.org/debian-user/2017/06/msg01196.html
> I have never been particularly happy with the solution I linked to in
> that thread, precisely because I was worried it would be damaged by
> updates, but it seems to be intact at this point, so I don't think that
> is what the problem is here]
>
> Thanks in advance if anyone can help me diagnose the sudden loss of the
> a2dp-sink protocol. The headphones work fine with my iPhone, so the
> problem seems to be with the computer, with which they were working fine
> until a bit over a week ago.
>
> Mark
>

Pulseaudio? If so what does pavucontrol say

Did Bluez updates come through? maybe pulseaudio-bluetooth module?

regards...


Of course, pavucontrol says nothing, because the headphones have not
been able to connect yet. As soon as they connect, the adp_sink error
disconnects them again, as I described in my original post.

The problem, I've got to think, is upstream of pulseaudio, no?

I don't know if bluez updates came through. There is no
pulseaudio-bluetooth package in Stretch -- there is a


>pulseaudio-module-bluetooth package, which is installed.


Just to check, I've just done another apt update and then apt list
--upgradeable to have a look at what would get installed now -- most of
it is to do with the freerdp security update from earlier this week,
nothing relating to pulse or bluetooth.

Thanks again

Mark


Mark,

What BT chipset are you using, have you acccidently disabled it in BIOS? Nothing shows up in google about protocol not 
available but there was a bug in gnome 3.24 and pulseaudio 10 with gnome/GDM where GDM starts its own pulseaudio 
instance. https://wiki.archlinux.org/index.php/Bluetooth_headset#Gnome_with_GDM


I don't run gnome on debian but i do on arch, i didn't find the gnome BT stack particularily stable so i used 
blueman-applet as a backup in the cases gnome BT wouldn't work with my FSL360. see blueman & bluez-utils packages in 
stretch.


Blueman-applet allows you see if the device is trusted & paired and to perform those functions if you wish. It also 
identifies the BT adapter in use which should appear in lsusb.


If that's no further help look up the Arch wiki section on bluetoothctl [from bluez-utils]. 99% of the time the issues 
i had were BT getting confused about adapters.


signature.asc
Description: PGP signature

Re: Live recording

[Doug]

On 08/03/2017 08:19 AM, Rodolfo Medina wrote:

Jeremy Nicoll  writes:


On Thu, 3 Aug 2017, at 13:23, Rodolfo Medina wrote:


3) mic2 is stereo.

Not necessarily.  It might be a mono signal that's been duplicated across two
outputs.  It's only 2-channel if it can record two separate sounds at the
same time.

You can get microphones which, in a single unit, record more than one
channel at once.  Their labelling or artwork on them normally makes it
clear that they have L and R (or more than that) inputs.

What makes and models are your mics?


I can't tell, because they're old and I don't keep their specs.  One of them
was given to me about 15 years ago, the other I don't even remember how I have
it.  But I've got another teo: they are, each of them, part of a headphone.
They also let hear the recorded sound from both channels.  They are:

  http://www.trust.com/it/product/11916-primo-chat-headset

and

  http://www.trust.com/it/product/20685-mauro-headset

Rodolfo


I looked at both of the links. Note that there are two connections for 
each device: headset (earphone) connection and mic connection. The 
headset connection will be
stereo, and the mic connection will be mono. If there is talkback (i.e., 
you can hear the mic output on the headphones,) the mic sound will 
appear in both earpieces,

but will not be stereo--it will just be the same sound in both earpieces.

You can't get something for nothing. If you want stereo, you will have 
to have two mics or a mic with two separate microphone elements aimed in 
two (left and right)
directions. Such a device will have a plug with three connections on it: 
left, right, and ground. Your PC will not be able to handle such a 
microphone! You would need
a stereo preamplifier to plug that mic into, and then you would plug the 
stereo output of the preamp into the LINE IN jack of the PC.


I would hope that this information will settle the question!

--doug

Re: Network config

[Henning Follmann]

On Thu, Aug 03, 2017 at 08:53:06PM +0200, Pascal Hambourg wrote:
> Le 03/08/2017 à 17:20, Joshua Schaeffer a écrit :
> > 
> > Configuration in /etc/network/interfaces only works when NetworkManager 
> > isn't installed,
> 
> Wrong. The default NetworkManager behaviour is not to manage interfaces
> defined in /etc/network/interfaces.
> 


Well, that is also partly wrong.
There was definitely an issue with a mix of /e/n/i and NM and systemd.
If the device couldn't be brought up during system start systemd would hang
sometimes for a very long time. This was at least true for Jessie.

-H

-- 
Henning Follmann   | hfollm...@itcfollmann.com

cups: Backend /usr/lib/cups/backend/usb does not exist! [solved]

[Simon Brandmair]

Hi everybody,

it took me quite a while to figure this out so I thought I post this
here for someone else in need to find this.

I wanted to reinstall my printer Brother HL-3040CN after I upgraded to
stretch. That didn't work at first. The installer failed with "Backend
/usr/lib/cups/backend/usb does not exist!". ldadmin gave 'Bad device-uri
scheme "usb"'.

The solution was easy, if I had known what to do: copy
/usr/lib/cups/backend-available/usb to /usr/lib/cups/backend/. Doi!

Maybe this will help someone in the future.

Simon

Re: Hébergeur à conseiller (pro-debian ?)

[Erwan David]

Le 08/03/17 à 20:17, Dominique Asselineau a écrit :
> kaliderus wrote on Thu, Aug 03, 2017 at 06:56:25PM +0200
>> Bonjour,
>>
>> Question simple :
>>
>> Quel hébergeur me conseillez-vous (ou déconseillez) et pourquoi ?
>>
>> Besoins :
>> - 1 nom de domaine
>> - 1 machine sur laquelle j'ai totalement la main, à priori physique,
>> donc rien de mutualisé
>> - échanges illimités
>> C'est un cadre professionnel, donc si on peut payer par chèque ou
>> virement chaque moi ça m'arrange (pas de carte de paiement :-( )
> 
> Il y a déjà ProxGroup qui répond à cette demande.  Toutefois, pas sûr
> qu'ils proposent des machines physiques mais sur leurs VPS, on a
> totalement la main.  Ils acceptent par exemple de configurer le
> reverse DNS, ce qui permet de gérer son propre serveur de messagerie.
> Pas sûr non plus qu'ils proposent un trafic illimité.  Ceci dit un
> trafic illimité est totalement illusoire puisque matériellement la
> bande passante est forcément limitée par les dures lois de la
> physique.  Les VPS sont rarement redémarrés.
> 
> 
> à propos des paiement par chèques ou autre virement, la périodicité
> est variable, du mois à l'année.
> 
> Il va sûrement y avoir d'autres suggestions.
> 

Sinon dedibox (online) je viens de vérifier, ils prennent les virement
("Paiements SEPA")

Re: Virtualisation de très vieux noyaux

[François TOURDE]

Le 17381ième jour après Epoch,
Olivier écrivait:

> Bonjour,
>
> J'ai une connaissance dont l'entreprise fait tourner des applications web
> métiers sur de très vieux noyaux datant de 2002, pour situer les choses
>
> Comme les machines qui hébergent ces applications tombent progressivement
> en panne, ma connaissance souhaite les remplacer par des machines
> virtuelles le temps que le remplacement des applications soit décidé, (ce
> temps là pouvant lui aussi se mesurer en années).

Une petite question: Les "applications web" en question ne tournent pas
sur des noyaux récents, j'imagine... Ou bien ne tournent pas sur des
distrib récentes ?

J'utilise Xen depuis de longues années, avec des VM qui boutent sur de
vieux noyaux. La condition semble être que le paquet du vieux noyau
ainsi que ses modules doivent être installés (et pas forcément utilisés)
sur le système hôte.

Re: Network config

[Pascal Hambourg]

Le 03/08/2017 à 17:20, Joshua Schaeffer a écrit :


Configuration in /etc/network/interfaces only works when NetworkManager isn't 
installed,


Wrong. The default NetworkManager behaviour is not to manage interfaces 
defined in /etc/network/interfaces.

Re: Network config

[Pascal Hambourg]

Le 03/08/2017 à 15:52, Zenaan Harkness a écrit :

On Thu, Aug 03, 2017 at 08:53:27AM -0400, Greg Wooledge wrote:

But the problem is, various Unix DHCP client daemons do *too much*.
All I want them to do is set the IP address, netmask, and gateway.
I *don't* want them to change the system hostname, or the system
resolv.conf (in which I have hand-placed *our* DNS search domain and
*our* DNS resolvers).


Well, making /etc/resolv.conf read-only, owned by root.root


... is just useless. resolv.conf is already owned by root, DCHP client 
daemons run as root and on Linux systems root (uid 0) ignores read/write 
permissions.



uninstalling resolvconf should also solve your problem.


No, it solves nothing. Without resolvonf the DHCP client will write 
directly in resolv.conf.



If you in this last (laptop) scenario need specific addition of your
static nameserver, on top of the DHCP nameservers (e.g. one
nameserver might resolve some internal names, the others might
recursively resolve internet names)


This does not work.
The resolver stops as soon as it receives a positive (record exists) or 
negative (record does not exist) answer. If it receives a negative 
answer from the first name server, it won't query the next name server.


All name servers declared in resolv.conf must be equal and provide the 
same answers, or unexpected behaviour will happen. Multiple name servers 
is only for redundancy and load balancing, not to provide different answers.

Re: Where to install Flash Player?

[Cindy-Sue Causey]

On 8/3/17, Thomas George  wrote:
> I downloaded flash-player-ppapi_26.0.0.137-1_amd64.deb into /opt and
> used dpkg -i to install it.
>
> Is this the best way to do it? dpkg -s flash-player-ppapi responds ok
> installed but dpkg -p flash-player-ppapi doesn't recognize it.


I'm not sure what to tell you on the "where", but I tried your "dpkg
-p" on several packages I have installed. I got the same message each
time:

dpkg-query: package 'inkscape' is not available
Use dpkg --info (= dpkg-deb --info) to examine archive files,
and dpkg --contents (= dpkg-deb --contents) to list their contents.

So I hit up "man dpkg". It shows "dpkg-query -p" as an option. Ahh, so
that's what the above might have been very subtly hinting. So I tried
that while not knowing what the original "dpkg -p" query returns. I'm
thinking this is probably not it:

dpkg-query: package 'inkscape' is not available
Use dpkg --info (= dpkg-deb --info) to examine archive files,
and dpkg --contents (= dpkg-deb --contents) to list their contents.

And I also tried both as root = same-same so I don't know. For what it
may yield in solutions, I did see this under "dpkg-query" (under "man
dpkg"):

"-p, --print-avail package-name...
  Display details about package-name, as found in
  /var/lib/dpkg/available. Users of APT-based frontends
  should use apt-cache show package-name instead."

Which I do personally use (that "apt-cache show package-name", I mean).

That's all on Buster :)

Cindy :)
-- 
Cindy-Sue Causey
Talking Rock, Pickens County, Georgia, USA

* runs with duct tape *

Re: KDE 5 usando mais memória do que o esperado

[Valentim Carlos]

Legal,



Vou dar uma olhada no blog sim e verificar esse post,



A principio eu  desabilitei  o Akonadi, ~/.config/akonadi/akonadiserverrc, 
alterei a linha StartServer = true para StartServer = false, o que resultou 
numa perda de aproximadamente 200K de memória.



É que tenho apache2, MySql e PHP rodando para desenvolvimento do Prestashop.



Mais não vou deixar de ver sobre desativar o "Compositor".



Agradeço novamente,



Valentim Carlos


Em 03 de ago de 2017 às 11:22, Wiliam Freitas  escreveu:


Uso KDE há anos e até hoje não me adaptei ao KDE 5 (pelo mesmo motivo
que vc comentou).

Desativando o "Compositor" você consegue algum ganho de performance e
economia de recursos. Fica utilizável, mas não tanto quanto o KDE 4.

Tenho um post com impressões sobre o KDE 5 no meu blog... aproveito para
divulgá-lo:

http://www.blogporta80.com.br/2015/08/03/artigo-kde-5-plasma-desktop-um-terrivel-desastre

Boa sorte!

[]'s

@wili4m




On 08/03/2017 09:42 AM, Valentim Carlos wrote:

Bom dia,


Estou com uma instalação nova do Debian com KDE 5 e esta
usando mais memória do que o esperado, Em Idle, sem nada aberto esta
usando por volta de 950 K, coisa que era pra ser entre 400 e 500 K, o
que nas minhas outras 3 maquinas estão usando. Com o Iceweasel aberto e
umas 3 páginas esta usando por volta de 2,2 GiB, o que nas outras
máquinas usa por volta de 1,6 GiB.
Alguma ideia sobre o que pode ocorrer isso ? pergunto
posi tenho que rodar Encoder e Render que usa muita memória ...


Agradeço antecipadamente,


Valentim Carlos

Re: Hébergeur à conseiller (pro-debian ?)

[Ph. Gras]

>> 
> 
> OVH acceptent-ils les chèques ?  C'était dans la demande initiale.

Oui, sous certaines conditions : les services sont activés seulement
lorsque le chèque est effectivement encaissé.

Ph. Gras

Créer un fichier grub.cfg correct

[andre_debian]

Comment recréer un fichier "grub.cfg" correct ?

Sur le disque dur, partition ext4 : sda2.
Deux sauvegardes ext4 sur un 2ème disque dur : sdb1 et sdb2.

# update-grub
me créé un fichier "grub.cfg" avec plein d'erreurs, contenant trop de 
lignes inutiles, avec des UUID ne correspondant pas aux partitions,
Exemple :
$menuentry_id_option 'gnulinux-simple-a8a150e7-77f0-4e27-bf93-33ab8932e427'
root=UUID=f9cafbc8-5b78-4b8a-9e5f-f50697ab33f4

Deux UUID différents pour la même partition !!!
Je veux booter sur sdb1 et me retrouve sur sda1.

Je désire un "grub.cfg" contenant par partition ceci :
un boot graphique + un boot recovery (single),
soit environ ~6 lignes,
avec toutes les lignes UUID qui correspondent bien aux 
/dev/sda(x) et sdb(x),
et supprimer ces moults boot à foison.

Y a t-il moyen de purger grub, car j'ai l'impression que
grub utilise d'anciennes données, soit repartir à zéro.

En supprimant grub.cfg, pareil, fichier pelin d'erreurs.

Merci, André

Re: Hébergeur à conseiller (pro-debian ?)

[Dominique Asselineau]

Ph. Gras wrote on Thu, Aug 03, 2017 at 08:14:41PM +0200
> Bonsoir,
> 
> >>> C'est un cadre professionnel, donc si on peut payer par chèque ou
> >>> virement chaque moi ça m'arrange (pas de carte de paiement :-( )
> 
> Gandi, OVH… ou n'importe quel autre si possible d'une capacité capitalistique
> équivalente à la boîte où tu bosses, pour éventuellement faire pression en cas
> de besoin. Si possible pas trop loin, histoire de pouvoir s'y déplacer ;-)
> 
> Gandi ne vend que des machines virtuelles onéreuses dès que tu commences
> à avaler du trafic. J'ai fait récemment basculer une assoce vers un Kimsufi, 
> car
> ce dernier revient aussi cher que les 2 disques qu'elle a chez Gandi, pour une
> capacité 3x plus importante.
> 
> OVH est une grosse boîte à présent, avec un service commercial dégueulasse.

OVH acceptent-ils les chèques ?  C'était dans la demande initiale.

Dominique

--

Re: Habilitar o KDE após instalação

[Wiliam Freitas]

Na tela de login deve haver uma opção (um menu) que lista todas 
interfaces instaladas no sistema.


Então basta selecionar o 'KDE' na hora do login.

[]'s

wili4m



On 07/28/2017 11:14 PM, Gilberto F da Silva wrote:

On Fri, Jul 28, 2017 at 10:07:08PM -0400, Rodolfo wrote:

Certo, você baixou o KDE, mas... instalou?


   Usei o apt-get install kde-full
   


# apt-get install kde-desktop


   Esse eu não instalei.  Vou voltar ao Debian e testar.
   

Re: KDE 5 usando mais memória do que o esperado

[Wiliam Freitas]

Uso KDE há anos e até hoje não me adaptei ao KDE 5 (pelo mesmo motivo 
que vc comentou).


Desativando o "Compositor" você consegue algum ganho de performance e 
economia de recursos. Fica utilizável, mas não tanto quanto o KDE 4.


Tenho um post com impressões sobre o KDE 5 no meu blog... aproveito para 
divulgá-lo:


http://www.blogporta80.com.br/2015/08/03/artigo-kde-5-plasma-desktop-um-terrivel-desastre

Boa sorte!

[]'s

@wili4m




On 08/03/2017 09:42 AM, Valentim Carlos wrote:

Bom dia,

   Estou com uma instalação nova do Debian com KDE 5 e esta 
usando mais memória do que o esperado, Em Idle, sem nada aberto esta 
usando por volta de 950 K, coisa que era pra ser entre 400 e 500 K, o 
que nas minhas outras 3 maquinas estão usando. Com o Iceweasel aberto e 
umas 3 páginas esta usando por volta de 2,2 GiB, o que nas outras 
máquinas usa por volta de 1,6 GiB.
Alguma ideia sobre o que pode ocorrer isso ? pergunto 
posi tenho que rodar Encoder e Render que usa muita memória ...


   Agradeço antecipadamente,

   Valentim Carlos

Re: Hébergeur à conseiller (pro-debian ?)

[Sandro CAZZANIGA]

> Le 3 août 2017 à 20:17, Sandro CAZZANIGA  a écrit 
> :
> 
> 
>> Le 3 août 2017 à 20:14, Ph. Gras  a écrit :
>> 
>> Bonsoir,
>> 
> C'est un cadre professionnel, donc si on peut payer par chèque ou
> virement chaque moi ça m'arrange (pas de carte de paiement :-( )
>> 
>> Gandi, OVH… ou n'importe quel autre si possible d'une capacité capitalistique
>> équivalente à la boîte où tu bosses, pour éventuellement faire pression en 
>> cas
>> de besoin. Si possible pas trop loin, histoire de pouvoir s'y déplacer ;-)
>> 
>> Gandi ne vend que des machines virtuelles onéreuses dès que tu commences
>> à avaler du trafic. J'ai fait récemment basculer une assoce vers un Kimsufi, 
>> car
>> ce dernier revient aussi cher que les 2 disques qu'elle a chez Gandi, pour 
>> une
>> capacité 3x plus importante.
>> 
>> OVH est une grosse boîte à présent, avec un service commercial dégueulasse.
>> 
>> À partir de 15-20 € par mois, tu as maintenant pléthore de fournisseurs…
>> 
>> Bonne continuation,
>> 
>> Ph. Gras
> 
> Bonsoir,
> 
> Sinon, tu as toujours la solution AWS, une petite instance EC2 (ou plus 
> grosse selon tes besoins).
> 
> Cordialement.
> 
> Sandro CAZZANIGA


Et j’oubliais, pour tes noms de domaines, AWS propose la solution « route 53 » 
pour déposer/transférer tes noms de domaine.

Cordialement.

Sandro CAZZANIGA



signature.asc
Description: Message signed with OpenPGP

Re: Live recording

[Rodolfo Medina]

Jeremy Nicoll  writes:

> On Thu, 3 Aug 2017, at 16:45, Rodolfo Medina wrote:
>
>> Thanks.  But all my 4 microphones terminates in 3.5mm, and the splitter I
>> was
>> talking about is two 3.5mm female and one 3.5mm male.  Is this you
>> mean...?
>
> No.  Look at: https://i.ytimg.com/vi/x1q9MXvjDlM/maxresdefault.jpg
>
> That should show you pictures of two, three and four contact mini jack
> plugs.
>
> A two-contact one is used for one audio channel (the two contacts being
> signal & ground).
>
> A three contact one is used for two channels (signal 1, signal 2, and
> ground).
>
> A four contact one is for three channels (eg on a stereo headset which
> has L & R speakers, 
> but a single mono microphone) - the four contacts being L, R, mic and
> ground.
>
> (The tip/ring/sleeve terms that someone else mentioned are names for the
> different contacts
> on a three-contact jack plug.) 
>
> Which type of plugs do you have on your mics?

All four my mics, all my headphones, all my adaptors, are all of type 2 in the
picture, where it says `stereo'.

> On 'professional' equipment it gets a bit more complicated when three
> connections are 
> used for a single audio channel, being the +ve half of a waveform, the
> -ve half, and ground.
> (These are known as 'balanced' connections, and typically use either XLR
> plugs & sockets
> or 3-contact 6.25mm audio jacks.)


Many thanks,

Rodolfo

Re: Hébergeur à conseiller (pro-debian ?)

[Sandro CAZZANIGA]

> Le 3 août 2017 à 20:14, Ph. Gras  a écrit :
> 
> Bonsoir,
> 
 C'est un cadre professionnel, donc si on peut payer par chèque ou
 virement chaque moi ça m'arrange (pas de carte de paiement :-( )
> 
> Gandi, OVH… ou n'importe quel autre si possible d'une capacité capitalistique
> équivalente à la boîte où tu bosses, pour éventuellement faire pression en cas
> de besoin. Si possible pas trop loin, histoire de pouvoir s'y déplacer ;-)
> 
> Gandi ne vend que des machines virtuelles onéreuses dès que tu commences
> à avaler du trafic. J'ai fait récemment basculer une assoce vers un Kimsufi, 
> car
> ce dernier revient aussi cher que les 2 disques qu'elle a chez Gandi, pour une
> capacité 3x plus importante.
> 
> OVH est une grosse boîte à présent, avec un service commercial dégueulasse.
> 
> À partir de 15-20 € par mois, tu as maintenant pléthore de fournisseurs…
> 
> Bonne continuation,
> 
> Ph. Gras

Bonsoir,

Sinon, tu as toujours la solution AWS, une petite instance EC2 (ou plus grosse 
selon tes besoins).

Cordialement.

Sandro CAZZANIGA


signature.asc
Description: Message signed with OpenPGP

Re: Hébergeur à conseiller (pro-debian ?)

[Dominique Asselineau]

kaliderus wrote on Thu, Aug 03, 2017 at 06:56:25PM +0200
> Bonjour,
> 
> Question simple :
> 
> Quel hébergeur me conseillez-vous (ou déconseillez) et pourquoi ?
> 
> Besoins :
> - 1 nom de domaine
> - 1 machine sur laquelle j'ai totalement la main, à priori physique,
> donc rien de mutualisé
> - échanges illimités
> C'est un cadre professionnel, donc si on peut payer par chèque ou
> virement chaque moi ça m'arrange (pas de carte de paiement :-( )

Il y a déjà ProxGroup qui répond à cette demande.  Toutefois, pas sûr
qu'ils proposent des machines physiques mais sur leurs VPS, on a
totalement la main.  Ils acceptent par exemple de configurer le
reverse DNS, ce qui permet de gérer son propre serveur de messagerie.
Pas sûr non plus qu'ils proposent un trafic illimité.  Ceci dit un
trafic illimité est totalement illusoire puisque matériellement la
bande passante est forcément limitée par les dures lois de la
physique.  Les VPS sont rarement redémarrés.


à propos des paiement par chèques ou autre virement, la périodicité
est variable, du mois à l'année.

Il va sûrement y avoir d'autres suggestions.

Dominique
--

Re: Hébergeur à conseiller (pro-debian ?)

[Ph. Gras]

Bonsoir,

>>> C'est un cadre professionnel, donc si on peut payer par chèque ou
>>> virement chaque moi ça m'arrange (pas de carte de paiement :-( )

Gandi, OVH… ou n'importe quel autre si possible d'une capacité capitalistique
équivalente à la boîte où tu bosses, pour éventuellement faire pression en cas
de besoin. Si possible pas trop loin, histoire de pouvoir s'y déplacer ;-)

Gandi ne vend que des machines virtuelles onéreuses dès que tu commences
à avaler du trafic. J'ai fait récemment basculer une assoce vers un Kimsufi, car
ce dernier revient aussi cher que les 2 disques qu'elle a chez Gandi, pour une
capacité 3x plus importante.

OVH est une grosse boîte à présent, avec un service commercial dégueulasse.

À partir de 15-20 € par mois, tu as maintenant pléthore de fournisseurs…

Bonne continuation,

Ph. Gras

Re: Live recording

[Rodolfo Medina]

David Wright  writes:

> On Thu 03 Aug 2017 at 15:43:15 (+0200), Rodolfo Medina wrote:
>> Rodolfo Medina  writes:
>> 
>> > Jeremy Nicoll  writes:
>> >
>> >> On Thu, 3 Aug 2017, at 13:23, Rodolfo Medina wrote:
>> >>
>> >>> 3) mic2 is stereo.
>> >>
>> >> Not necessarily.  It might be a mono signal that's been duplicated across
>> >> two outputs.
>> 
>> Is there, as far as you know, a mono microphone model that we can be sure it
>> does not do so?  I.e., its sound is heared on only one channel?
>
> Most mono microphones will be fitted with a plug that has a tip and
> barrel, but no ring (between the two). The effects you observe could
> be no more than chance vagarities in the mechanical construction of
> plug and socket.
>
>> >> It's only 2-channel if it can record two separate sounds at the
>> >> same time.
>> >>
>> >> You can get microphones which, in a single unit, record more than one 
>> >> channel at once.  Their labelling or artwork on them normally makes it 
>> >> clear that they have L and R (or more than that) inputs.
>> >>
>> >> What makes and models are your mics?
>> >
>> >
>> > I can't tell, because they're old and I don't keep their specs.  One of
>> > them was given to me about 15 years ago, the other I don't even remember
>> > how I have it.  But I've got another teo: they are, each of them, part of
>> > a headphone.  They also let hear the recorded sound from both channels.
>> > They are:
>> >
>> >  http://www.trust.com/it/product/11916-primo-chat-headset
>> >
>> > and
>> >
>> >  http://www.trust.com/it/product/20685-mauro-headset
>> 
>> 
>> Just now I received a reply from Trust assistence and they say the mic
>> (Primo) is mono.
>> 
>> Anyway, about my example above, can we conclude that mic1 is mono and that
>> the mic input of my PC is stereo...?
>
> If you want to know whether the input is stereo, plug in a cable which
> has a stereo 3.5mm plug going to two phono plugs. (These computer↔monitor
> sound cables were everywhere at one time.) Touch just the pin of each
> phono plug in turn and see if you get mains hum separately on the
> left and right channels of the computer.

I used a cable thath has, at one end, a 3.5mm stereo male plug; and, at the
other end, a couple of twin phono RCA males.  I plugged the 3.5mm into the mic
input of the PC.  Then I did:

 $ sox -t alsa default output.wav

Then I touched alternatively one and the other of the two phone plugs and saw
that the output of sox alternatively filled up the left or right channel.


> If you're keen to use these microphones and you have the budget and
> attitude of a school pupil, buy two mono 3.5mm sockets and a twin
> phono socket unit. Solder the connections yourself using the phone
> socket pair to anchor the 3.5mm ones and give it stability. Then
> connect this device to the computer with the aforementioned cable
> and plug in the mikes.

Not very clear to me...  But I'm not good at soldering...

Many thanks,

Rodolfo

Re: Flash Player Problem

[Patrick Bartek]

On Thu, 3 Aug 2017 13:38:37 -0400 Thomas George 
wrote:

> An attempt to log in to the Bridgebase website fails.
> 
> Login using firefox ESR results in an all blue screen. The firefox
> plugin shockwave flash is installed.
> 
> My system is debian stretch using the LXDE desktop. I have also
> installed gnash, gnash-common, and browser-plugin-gnash in an attempt
> to solve this problem.
> 
> Login using Chromium browser fails with note to install flash player
> BUT dpkg -s flash-player-ppapi reports installed ok.
> 
> bridgebase is the only website that has this problem. I have
> contacted their support
> and they suggest problem is changes in flash or in the browsers.
> 
> Can you help?

Do you have the most current  Flash installed? 26.0.0.137  To find out,
in Firefox type "about:plugins"  (without the quotes) in the URL field.

B

Unable to change mouse acceleration and threshold in Stretch

[Илья Валеев]

Hello!

I'm unable to change mouse acceleration and threshold in Stretch.

It does not depend on DE. There are no error appears. For example, in
KDE settings applies correctly, ~/.kde/share/config/kcminputrc creates
and contain properly values, but mouse behavior does not changes.
'xset m 3 1' also does not change anything.

It does not depend on:
- Stretch 9.0 or Stretch 9.1
- is it clean install or not
- mouse
- is it virtual machine with Jessie host or real system

Can I do something with this or should I create a bugreport?
If I should create a bugreport, what package should I specify?

P.S. Maybe another bug I've encountered associated with mouse bug
described above: in Dolphin mousewheel scrolls not the number of lines
specified in config but same amount of pixels.

Where to install Flash Player?

[Thomas George]

I downloaded flash-player-ppapi_26.0.0.137-1_amd64.deb into /opt and 
used dpkg -i to install it.


Is this the best way to do it? dpkg -s flash-player-ppapi responds ok 
installed but dpkg -p flash-player-ppapi doesn't recognize it.

Re: Hébergeur à conseiller (pro-debian ?)

[contact]

  
  
Le 03/08/2017 à 19:52, Olivier Lange a
  écrit :


  OVH?

Le 3 août 2017 à 18:56, kaliderus  a écrit :

  
Bonjour,

Question simple :

Quel hébergeur me conseillez-vous (ou déconseillez) et pourquoi ?

Besoins :
- 1 nom de domaine
- 1 machine sur laquelle j'ai totalement la main, à priori physique,
donc rien de mutualisé
- échanges illimités
C'est un cadre professionnel, donc si on peut payer par chèque ou
virement chaque moi ça m'arrange (pas de carte de paiement :-( )
di
Merci la liste.
K


  
  



ou gandi 

-- 
  François-Marie BILLARD
  Sculpteur - Céramiste 
  Atelier à Avaray - Loir-et-Cher.
  

Re: Flash Player Problem

[Curt]

On 2017-08-03, Thomas George  wrote:
> An attempt to log in to the Bridgebase website fails.

Seems to work here (Wheezy, Firefox-esr).

> Login using firefox ESR results in an all blue screen. The firefox
> plugin shockwave flash is installed.

Background is indeed glaucous, but in the foreground I have the login
prompt (it all comes up in a separate pop-up window).

> My system is debian stretch using the LXDE desktop. I have also
> installed gnash, gnash-common, and browser-plugin-gnash in an attempt
> to solve this problem.
>
> Login using Chromium browser fails with note to install flash player
> BUT dpkg -s flash-player-ppapi reports installed ok.
>
> bridgebase is the only website that has this problem. I have contacted 
> their support
> and they suggest problem is changes in flash or in the browsers.
>
> Can you help?
>

I can't help. (Clear cookies and cache and allow pop-ups for the bridgebase
site?)


-- 
“Certitude is not the test of certainty.”
--Oliver Wendell Holmes, Jr.

Re: Flash Player Problem

[Joe]

On Thu, 3 Aug 2017 13:38:37 -0400
Thomas George  wrote:

> An attempt to log in to the Bridgebase website fails.
> 
> Login using firefox ESR results in an all blue screen. The firefox
> plugin shockwave flash is installed.
> 
> My system is debian stretch using the LXDE desktop. I have also
> installed gnash, gnash-common, and browser-plugin-gnash in an attempt
> to solve this problem.
> 
> Login using Chromium browser fails with note to install flash player
> BUT dpkg -s flash-player-ppapi reports installed ok.
> 
> bridgebase is the only website that has this problem. I have
> contacted their support
> and they suggest problem is changes in flash or in the browsers.
> 
> Can you help?
> 

No, but I've seen the same kind of problem with the Halifax bank site
in the UK in the last couple of weeks. It crashes Firefox and the login
doesn't work with Konqueror, but fortunately does work with Midori.

No mention of flash anywhere, but the culprit is certain to be some
stupid marketing script on the page. I'm not going to buy anything if I
can't see it because my browser crashed...

-- 
Joe

Re: Hébergeur à conseiller (pro-debian ?)

[Olivier Lange]

OVH?

Le 3 août 2017 à 18:56, kaliderus  a écrit :
> Bonjour,
>
> Question simple :
>
> Quel hébergeur me conseillez-vous (ou déconseillez) et pourquoi ?
>
> Besoins :
> - 1 nom de domaine
> - 1 machine sur laquelle j'ai totalement la main, à priori physique,
> donc rien de mutualisé
> - échanges illimités
> C'est un cadre professionnel, donc si on peut payer par chèque ou
> virement chaque moi ça m'arrange (pas de carte de paiement :-( )
>
> Merci la liste.
> K
>

Flash Player Problem

[Thomas George]

An attempt to log in to the Bridgebase website fails.

Login using firefox ESR results in an all blue screen. The firefox
plugin shockwave flash is installed.

My system is debian stretch using the LXDE desktop. I have also
installed gnash, gnash-common, and browser-plugin-gnash in an attempt
to solve this problem.

Login using Chromium browser fails with note to install flash player
BUT dpkg -s flash-player-ppapi reports installed ok.

bridgebase is the only website that has this problem. I have contacted 
their support

and they suggest problem is changes in flash or in the browsers.

Can you help?

RE: Live recording

[larry owens]

FYI the "tip, ring and sleeve" notation derives from the 3-wire connection that 
connects old land-lines to the telephone central office in the user's  area.
Larry

-Original Message-
From: Jeremy Nicoll [mailto:jn.ml.dbn...@letterboxes.org] 
Sent: Thursday, August 03, 2017 10:07 AM
To: debian-user@lists.debian.org
Subject: Re: Live recording

On Thu, 3 Aug 2017, at 16:45, Rodolfo Medina wrote:

> Thanks.  But all my 4 microphones terminates in 3.5mm, and the splitter I
> was
> talking about is two 3.5mm female and one 3.5mm male.  Is this you
> mean...?

No.  Look at: https://i.ytimg.com/vi/x1q9MXvjDlM/maxresdefault.jpg

That should show you pictures of two, three and four contact mini jack
plugs.

A two-contact one is used for one audio channel (the two contacts being
signal & ground).

A three contact one is used for two channels (signal 1, signal 2, and
ground).

A four contact one is for three channels (eg on a stereo headset which
has L & R speakers, 
but a single mono microphone) - the four contacts being L, R, mic and
ground.

(The tip/ring/sleeve terms that someone else mentioned are names for the
different contacts
on a three-contact jack plug.) 

Which type of plugs do you have on your mics?


On 'professional' equipment it gets a bit more complicated when three
connections are 
used for a single audio channel, being the +ve half of a waveform, the
-ve half, and ground.
(These are known as 'balanced' connections, and typically use either XLR
plugs & sockets
or 3-contact 6.25mm audio jacks.)

-- 
Jeremy Nicoll - my opinions are my own.

Java perspective missing in Eclipse 3.8.1-10

[rudu]

Hello,

It was almost a year ago when I last used Eclipse to work on some Java 
projects and everything was going fine.

Today, it freezed on its splash screen on every launch.
I first removed ~/.eclipse to no avail.
When I removed path/to/my/workspace/.metadata Eclipse did open a fresh 
window where I should have been able to re-create my Java project ... 
unless there was no way to initiate Java projects anymore.

I purge eclipse and its dependencies :
# apt-get autoremove eclipse* --purge

Then did a fresh install :
$ dpkg -l eclipse* | grep ^ii
ii  eclipse 3.8.1-10 all  Extensible Tool 
Platform and Java IDE
ii  eclipse-jdt 3.8.1-10 all  Eclipse Java 
Development Tools (JDT)
ii  eclipse-pde 3.8.1-10 amd64Eclipse Plug-in 
Development Environment (PDE)
ii  eclipse-platform3.8.1-10 amd64Eclipse platform 
without development plug-ins
ii  eclipse-platform-data   3.8.1-10 all  Eclipse platform 
without development plug-ins (data)
ii  eclipse-rcp 3.8.1-10 amd64Eclipse Rich 
Client Platform (RCP)


Still, there is no way to generate a java project.
Did I overlooked anything ?

I'am running debian testing.

Thanks in advance for any hint.

Rudu

Re: Live recording

[Jeremy Nicoll]

On Thu, 3 Aug 2017, at 16:45, Rodolfo Medina wrote:

> Thanks.  But all my 4 microphones terminates in 3.5mm, and the splitter I
> was
> talking about is two 3.5mm female and one 3.5mm male.  Is this you
> mean...?

No.  Look at: https://i.ytimg.com/vi/x1q9MXvjDlM/maxresdefault.jpg

That should show you pictures of two, three and four contact mini jack
plugs.

A two-contact one is used for one audio channel (the two contacts being
signal & ground).

A three contact one is used for two channels (signal 1, signal 2, and
ground).

A four contact one is for three channels (eg on a stereo headset which
has L & R speakers, 
but a single mono microphone) - the four contacts being L, R, mic and
ground.

(The tip/ring/sleeve terms that someone else mentioned are names for the
different contacts
on a three-contact jack plug.) 

Which type of plugs do you have on your mics?


On 'professional' equipment it gets a bit more complicated when three
connections are 
used for a single audio channel, being the +ve half of a waveform, the
-ve half, and ground.
(These are known as 'balanced' connections, and typically use either XLR
plugs & sockets
or 3-contact 6.25mm audio jacks.)

-- 
Jeremy Nicoll - my opinions are my own.

Hébergeur à conseiller (pro-debian ?)

[kaliderus]

Bonjour,

Question simple :

Quel hébergeur me conseillez-vous (ou déconseillez) et pourquoi ?

Besoins :
- 1 nom de domaine
- 1 machine sur laquelle j'ai totalement la main, à priori physique,
donc rien de mutualisé
- échanges illimités
C'est un cadre professionnel, donc si on peut payer par chèque ou
virement chaque moi ça m'arrange (pas de carte de paiement :-( )

Merci la liste.
K

Re: Fonts readability (was: Arial vs. Helvetica.)

[Henrique de Moraes Holschuh]

On Thu, 03 Aug 2017, Nicolas George wrote:
> Another point where the bitmap fonts beat the vectorial fonts at tiny
> sizes: you usually want your vectorial fonts anti-aliased, but at tiny
> sizes it hurts readability. Even worse, the anti-aliasing is done wrong:
> it is done without taking gamma correction into account. That means that
> when 50% intensity is wanted, it produces 22% intensity instead:
> black-on-white is too thick, white-on-black is too thin.

Professional, high-quality fonts *optimized for small sizes* will have
specific glyph variations and "rendering rules" for the small sizes, and
do not suffer from the annoying misplaced anti-aliasing effects (i.e.
they will render about as nicely as a pixel-optimized bitmapped font).

However that also requires that the font rendering engine do it
perfectly (which in turn requires it to implement everything required,
for whatever is using it to *enable* all of that and not screw it up
when blitting the result, etc).

So, I am not at all surprised that it is far easier to get better
results for small sizes and terminal fonts using bitmapped fonts :-)  It
should be much faster for the terminal to use the bitmapped ones, too.

-- 
  Henrique Holschuh

Re: Live recording

[Rodolfo Medina]

David Wright  writes:

> On Thu 03 Aug 2017 at 15:43:15 (+0200), Rodolfo Medina wrote:
>> Rodolfo Medina  writes:
>> 
>> > Jeremy Nicoll  writes:
>> >
>> >> On Thu, 3 Aug 2017, at 13:23, Rodolfo Medina wrote:
>> >>
>> >>> 3) mic2 is stereo.
>> >>
>> >> Not necessarily.  It might be a mono signal that's been duplicated across
>> >> two outputs.
>> 
>> Is there, as far as you know, a mono microphone model that we can be sure it
>> does not do so?  I.e., its sound is heared on only one channel?
>
> Most mono microphones will be fitted with a plug that has a tip and
> barrel, but no ring (between the two). The effects you observe could
> be no more than chance vagarities in the mechanical construction of
> plug and socket.
>
>> >> It's only 2-channel if it can record two separate sounds at the
>> >> same time.
>> >>
>> >> You can get microphones which, in a single unit, record more than one 
>> >> channel at once.  Their labelling or artwork on them normally makes it 
>> >> clear that they have L and R (or more than that) inputs.
>> >>
>> >> What makes and models are your mics?
>> >
>> >
>> > I can't tell, because they're old and I don't keep their specs.  One of
>> > them was given to me about 15 years ago, the other I don't even remember
>> > how I have it.  But I've got another teo: they are, each of them, part of
>> > a headphone.  They also let hear the recorded sound from both channels.
>> > They are:
>> >
>> >  http://www.trust.com/it/product/11916-primo-chat-headset
>> >
>> > and
>> >
>> >  http://www.trust.com/it/product/20685-mauro-headset
>> 
>> 
>> Just now I received a reply from Trust assistence and they say the mic
>> (Primo) is mono.
>> 
>> Anyway, about my example above, can we conclude that mic1 is mono and that
>> the mic input of my PC is stereo...?
>
> If you want to know whether the input is stereo, plug in a cable which
> has a stereo 3.5mm plug going to two phono plugs. (These computer↔monitor
> sound cables were everywhere at one time.) Touch just the pin of each
> phono plug in turn and see if you get mains hum separately on the
> left and right channels of the computer.
>
> If you're keen to use these microphones and you have the budget and
> attitude of a school pupil, buy two mono 3.5mm sockets and a twin
> phono socket unit. Solder the connections yourself using the phone
> socket pair to anchor the 3.5mm ones and give it stability. Then
> connect this device to the computer with the aforementioned cable
> and plug in the mikes.


Thanks.  But all my 4 microphones terminates in 3.5mm, and the splitter I was
talking about is two 3.5mm female and one 3.5mm male.  Is this you mean...?

Rodolfo

Re: Fonts readability (was: Arial vs. Helvetica.)

[Nicolas George]

Le sextidi 16 thermidor, an CCXXV, rhkra...@gmail.com a écrit :
> > Even worse, the anti-aliasing is done wrong:
> > it is done without taking gamma correction into account. That means that
> > when 50% intensity is wanted, it produces 22% intensity instead:
> > black-on-white is too thick, white-on-black is too thin.
> 
> Thank you for confirming something I suspected for a long time.
> 
> Do you know what program(s) is responsibe for the anti-aliasing,
> against which a bug might be filed (or maybe a bug has already been
> filed)?

It is hard to tell, there are several components working together, and
they all will try to shift the blame to each other. I tried filling it
years ago, and it gave me that:

http://bugs.freedesktop.org/show_bug.cgi?id=13431

If somebody dares tell me that the way of reading text that I chose is
"wrong", I think it is worthless to try to discuss.

The components in play are:

- fontconfig: at some point, the user needs to be able to tell that
  gamma correction must be taken into account and what the gamma value
  should be. Despite what kp wrote, it belongs in fontconfig, just as
  much as lcdfilter or rgba.

- Freetype: it is the component that rasterizes the vector fonts,
  including anti-aliasing, so it would be an easy place to adjust for
  gamma. Unfortunately, Freetype does not know the color of the text and
  background, its output could be considered an alpha map.

- The X11 RENDER extension: it is the component that performs the alpha
  blending. It would be the correct place to implement the gamma
  adjustment. Unfortunately, it has no provisions to do so.

- Xft: it is the glue that holds everything together. I do not think it
  needs any work for this issue.

Good luck if you want to get things moving.

Regards,

-- 
  Nicolas George


signature.asc
Description: Digital signature

Re: Network config

[Joshua Schaeffer]

On 08/02/2017 06:56 PM, Zenaan Harkness wrote:
>
> I've preferred a static networking config for years, and resolvconf
> works well in this situation - but once resolvconf is configured,
> I've always put the dns setting in /etc/networks/interfaces

I agree with this as well. If you want to use static configuration and want 
resolvconf installed, you need to use the dns-* options in 
/etc/network/interfaces. resolvconf will then update your /etc/resolv.conf file 
with those options. If you change /etc/resolv.conf manually, resolvconf will 
override those settings periodically. The alternative is to simply uninstall 
resolvconf and set your DNS settings manually.

>  - the
> only time I put an entry in /etc/resolv.conf is when I'm testing
> stuff or doing a quick hack.
>
> Static network configs are quicker and give that sense of control -
> if the gui is down I can still fix things, and my knowledge applies
> in both gui and console scanerios.

Configuration in /etc/network/interfaces only works when NetworkManager isn't 
installed, which it typically is in GUI environments. If you don't have it 
installed in your GUI environment then yes, it works in both. If NetworkManager 
is installed then the nmcli command should be used and you shouldn't do any 
configuration in /etc/network/interfaces (although loopback is typically still 
controlled through this file).

Thanks,
Joshua Schaeffer

Thinkpad function keys in Stretch

[Proxy]

Hello,

After update to Debian Stretch, I'm having problem with function keys on
T440s. On Jessie, everything worked out of the box, but now volume up
and down, screen brightness and mic mute don't work. Sound mute work,
but only partialy. Led indicator doesn't turns on and it just mute the
sound, it doesn't toggle it, like before. 

$ cat /proc/acpi/ibm/volume 
level:  unsupported
mute:   off


thinkpad_acpi is loaded. 

lsmod |grep thinkpad
thinkpad_acpi  86016  1
nvram  16384  1 thinkpad_acpi
snd86016  15 
snd_hda_intel,snd_hwdep,snd_hda_codec,snd_timer,thinkpad_acpi,snd_hda_codec_hdmi,snd_hda_codec_generic,snd_hda_codec_realtek,snd_pcm
rfkill 24576  7 bluetooth,thinkpad_acpi,cfg80211
video  40960  2 thinkpad_acpi,i915

Here is dmesg:

dmesg | grep thinkpad 
[   10.656995] thinkpad_acpi: ThinkPad ACPI Extras v0.25
[   10.656996] thinkpad_acpi: http://ibm-acpi.sf.net/
[   10.656997] thinkpad_acpi: ThinkPad BIOS GJET82WW (2.32 ), EC unknown
[   10.656997] thinkpad_acpi: Lenovo ThinkPad T440s, model 20ARS08L01
[   10.657606] thinkpad_acpi: radio switch found; radios are enabled
[   10.657708] thinkpad_acpi: This ThinkPad has standard ACPI backlight 
brightness control, supported by the ACPI video driver
[   10.657709] thinkpad_acpi: Disabling thinkpad-acpi brightness events by 
default...
[   10.659307] thinkpad_acpi: rfkill switch tpacpi_bluetooth_sw: radio is 
unblocked
[   10.660120] thinkpad_acpi: rfkill switch tpacpi_wwan_sw: radio is unblocked
[   10.665657] input: ThinkPad Extra Buttons as 
/devices/platform/thinkpad_acpi/input/input8


So, it seams that thinkpad_acpi disables the brightnes control.

xbacklight doesn't work either. 

$ xbacklight -get
No outputs have backlight property

$ ls /sys/class/backlight/intel_backlight/
actual_brightness  bl_power  brightness  device  max_brightness  power 
subsystem  type  uevent


I tried various kernel options:

acpi_osi="Linux"
thinkpad-acpi.brightness_enable=1
acpi_osi="!Windows\ 2012"
video.use_native_backlight=1
acpi_backlight=vendor
acpi_backlight=video
acpi_backlight=native
acpi_backlight=none


in all combinations, without success. 

I also tried linux-image-4.11.0-0.bpo.1-amd64 kernel, but still no joy. 

Tried old kernel, 3.16.0-4, too, but this time it didn't work either.

I'm using 4.9.0-3 at the moment.

Any idea what could be wrong?


Thanks!

Re: Virtualisation de très vieux noyaux

[Timote Brusson]

Bonjour,
personnellement j'utilise l'outil de Vmware pour capturer des systèmes
vieillissants et les déployer sur un EXSi.
Avec vCenter Converter aucuns soucis.
Je l'ai déjà fait plusieurs fois avec des RedHat 2.1 et des Debian 3.
Bien cordialement,

-- 
Brusson Timoté

Re: Live recording

[David Wright]

On Thu 03 Aug 2017 at 15:43:15 (+0200), Rodolfo Medina wrote:
> Rodolfo Medina  writes:
> 
> > Jeremy Nicoll  writes:
> >
> >> On Thu, 3 Aug 2017, at 13:23, Rodolfo Medina wrote:
> >>
> >>> 3) mic2 is stereo.
> >>
> >> Not necessarily.  It might be a mono signal that's been duplicated across
> >> two outputs.
> 
> Is there, as far as you know, a mono microphone model that we can be sure it
> does not do so?  I.e., its sound is heared on only one channel?

Most mono microphones will be fitted with a plug that has a tip and
barrel, but no ring (between the two). The effects you observe could
be no more than chance vagarities in the mechanical construction of
plug and socket.

> >> It's only 2-channel if it can record two separate sounds at the
> >> same time.
> >>
> >> You can get microphones which, in a single unit, record more than one 
> >> channel at once.  Their labelling or artwork on them normally makes it 
> >> clear that they have L and R (or more than that) inputs.
> >>
> >> What makes and models are your mics?
> >
> >
> > I can't tell, because they're old and I don't keep their specs.  One of them
> > was given to me about 15 years ago, the other I don't even remember how I
> > have it.  But I've got another teo: they are, each of them, part of a
> > headphone.  They also let hear the recorded sound from both channels.  They
> > are:
> >
> >  http://www.trust.com/it/product/11916-primo-chat-headset
> >
> > and
> >
> >  http://www.trust.com/it/product/20685-mauro-headset
> 
> 
> Just now I received a reply from Trust assistence and they say the mic (Primo)
> is mono.
> 
> Anyway, about my example above, can we conclude that mic1 is mono and that the
> mic input of my PC is stereo...?

If you want to know whether the input is stereo, plug in a cable which
has a stereo 3.5mm plug going to two phono plugs. (These computer↔monitor
sound cables were everywhere at one time.) Touch just the pin of each
phono plug in turn and see if you get mains hum separately on the
left and right channels of the computer.

If you're keen to use these microphones and you have the budget and
attitude of a school pupil, buy two mono 3.5mm sockets and a twin
phono socket unit. Solder the connections yourself using the phone
socket pair to anchor the 3.5mm ones and give it stability. Then
connect this device to the computer with the aforementioned cable
and plug in the mikes.

Cheers,
David.

Re: Redmine on server in DMZ

[Jörg-Volker Peetz]

Johann Spies wrote on 08/02/17 11:56:
> A few years ago we got redmine working on a server in the DMZ on our
> network using ruby gems copied there from another computer with access
> to the internet.
> 
> This server has now been updated to Debian Stretch - which broke the
> redmine installation.
> 
> None of us know Ruby-on-Rails or Ruby well enough to debug the errors
> or to rectify what is wrong at the moment.
> 
> I tested a new redmine-pgsql installation on another computer with
> free access to the internet and it seems that as part of the post
> installation process it installed a lot of gems.  We cannot do that in
> the DMZ because the server cannot reach out to the internet.
> 
I have a redmine server running on testing. I'm not aware of such network
accesses except for the debian packages.
How did you install the new one, "aptitude install redmine"?
Could you post the output of the installation command?

> Is it possible to install redmine without having to install extra gems?
> 
> Regards
> Johann
> 
Regards,
jvp.

Re: Virtualisation de très vieux noyaux

[Olivier]

Merci André: ton témoignage est vraiment très, très intéressant !

Le 3 août 2017 à 15:32, Andre Majorel  a écrit :

> On 2017-08-03 15:09 +0200, Olivier wrote:
>
> > J'ai une connaissance dont l'entreprise fait tourner des
> > applications web métiers sur de très vieux noyaux datant de
> > 2002, pour situer les choses
> >
> > Comme les machines qui hébergent ces applications tombent
> > progressivement en panne, ma connaissance souhaite les
> > remplacer par des machines virtuelles le temps que le
> > remplacement des applications soit décidé, (ce temps là
> > pouvant lui aussi se mesurer en années).
> >
> > Pour être le plus clair possible, est-il possible de copier un
> > disque d'une très vieille machine (disons Debian 2.1 Slink)
> > sur un disque d'une VM quelconque (ESXi, KVM, ...) et espérer
> > que ça marche pendant encore 1 an ou
> > 2 ?
> >
> > Avez-vous des retours d'expérience réussie ou non, sur ce type
> > d'opération ?
>
> À cette époque, j'installais dans des VMware toutes sortes
> d'Unix a priori pas prévus pour la virtualisation. Ici, c'est
> pareil sauf qu'il ne s'agit pas d'une nouvelle installation mais
> d'une migration.
>
> Le difficultés viendront des périphériques. En particulier le
> réseau : il faut que la VM sache émuler une carte ethernet pour
> laquelle il existe un driver dans le kernel. De préférence la
> même que celle installée dans le PC sur lequel l'OS a été
> installé à l'origine.
>
> --
> André Majorel 
> Ever got spam through an address harvested from lists.debian.org ?
> Neither have I.
>
>

Re: How to Work Around Bug?

[Kent West]

On Wed, Aug 2, 2017 at 4:15 PM, Sven Joachim  wrote:

> On 2017-08-02 15:51 -0500, Kent West wrote:
>
> > I'm running Sid (so I know brokenness happens), and I got bit by bug
> > #869670, "linux-headers-4.11.0-2-amd64 : Depends:
> > linux-headers-4.11.0-2-common (= 4.11.11-1+b1) but it is not going to be
> > installed".
> >
> > I expected that a "grave" bug might get fixed pretty rapidly, so I've not
> > worried too much about it, refreshing the bug report screen in my
> > web-browser two or three times a day looking for any movement, but it's
> > been nearly two weeks, and the only movement I've seen was me chiming in
> a
> > "Me too, updates?" type post.
>
> It's fixed in the git master branch, but this branch currently tracks
> Linux 4.12, no idea when that might hit unstable.
>
> > So I'm guessing it's not going to be fixed anytime soon. Can someone
> > suggest a work-around? Perhaps a switch to aptitude/apt-get for force the
> > install despite the broken dependency, etc? I need the header files for
> > VirtualBox modules to recompile, so I can once again run my VirtualBox
> > machines.
>
> You can obtain the previous version of linux-headers-4.11.0-2-amd64 from
> snapshot.debian.org:
>
> http://snapshot.debian.org/package/linux/4.11.11-1/#
> linux-headers-4.11.0-2-amd64_4.11.11-1
>
> Cheers,
>Sven
>
>
"dpkg -i [that package]" (along with a little tinkering) got my VirtualBox
back up and running.

Thank you!

-- 
Kent West<")))><
Westing Peacefully - http://kentwest.blogspot.com

Re: Sudden death of bluetooth headphone connectivity

[Mark Fletcher]

On Thu, Aug 03, 2017 at 02:03:36PM +0100, stuart watt wrote:
> On 08/03, Mark Fletcher wrote:
> > Hello the list!
> > 
> > Suddenly, earlier this week, my bluetooth headphones stopped working
> > with Stretch.
> > 
> > I update weekly, usually on Sundays, and I am not sure because I wasn't
> > paying attention whether I had successfully used my headphones before
> > the last update I did last Sunday.
> > 
> > They definitely were working properly one week before that, when I used
> > them with the computer just before going off on a week-long business
> > trip.
> > 
> > Anyway, I use Gnome. When I want to connect the headphones I turn them
> > on, click on the network-like icon in the top right of the screen, click
> > on "Not in Use" beside the Bluetooth icon, and then click on "Bluetooth
> > Settings". (Is there a more efficient, involving fewer clicks, way to
> > connect, by the way?)
> > 
> > In Bluetooth Settings I can see my headphones as "BeoPlay H8". If I
> > click on them it brings up the window with the slider switch to connect
> > them. Normally, what would have happened is that clicking that switch
> > would have connected the headphones. WHat happens now is the switch goes
> > very briefly to the ON position, and then goes straight back to the OFF
> > position. Removing / forgetting the device and re-pairing leads to the
> > same result -- pairing is successful, but cannot connect.
> > 
> > With journalctl -f running in a terminal while trying to connect I see
> > the following at the moment of clicking the slider switch to connect the
> > headphones: a2dp-sink profile connect failed for AA:BB:CC:DD:EE:FF:
> > Protocol not available
> >  > address>
> > 
> > Now, call me a super-sleuth, but I suspect that Protocol Not Available
> > error could be something to do with the problem ;)
> > 
> > To re-iterate, this was working until recently, and all I've done is
> > install regular updates using apt update / apt upgrade.
> > 
> > Where should I start looking?
> > 
> > [Full disclosure -- I previously had a different problem on upgrading to
> > Stretch from Jessie -- see this link for that problem and a link to its
> > solution --> https://lists.debian.org/debian-user/2017/06/msg01196.html
> > I have never been particularly happy with the solution I linked to in
> > that thread, precisely because I was worried it would be damaged by
> > updates, but it seems to be intact at this point, so I don't think that
> > is what the problem is here]
> > 
> > Thanks in advance if anyone can help me diagnose the sudden loss of the
> > a2dp-sink protocol. The headphones work fine with my iPhone, so the
> > problem seems to be with the computer, with which they were working fine
> > until a bit over a week ago.
> > 
> > Mark
> > 
> 
> Pulseaudio? If so what does pavucontrol say
> 
> Did Bluez updates come through? maybe pulseaudio-bluetooth module?
> 
> regards...

Of course, pavucontrol says nothing, because the headphones have not 
been able to connect yet. As soon as they connect, the adp_sink error 
disconnects them again, as I described in my original post.

The problem, I've got to think, is upstream of pulseaudio, no?

I don't know if bluez updates came through. There is no 
pulseaudio-bluetooth package in Stretch -- there is a 
pulseaudio-module-bluetooth package, which is installed.

Just to check, I've just done another apt update and then apt list 
--upgradeable to have a look at what would get installed now -- most of 
it is to do with the freerdp security update from earlier this week, 
nothing relating to pulse or bluetooth. 

Thanks again

Mark

Re: Network config

[Zenaan Harkness]

On Thu, Aug 03, 2017 at 08:53:27AM -0400, Greg Wooledge wrote:
> But the problem is, various Unix DHCP client daemons do *too much*.
> All I want them to do is set the IP address, netmask, and gateway.
> I *don't* want them to change the system hostname, or the system
> resolv.conf (in which I have hand-placed *our* DNS search domain and
> *our* DNS resolvers).

Well, making /etc/resolv.conf read-only, owned by root.root, and
uninstalling resolvconf should also solve your problem.

Then programs won't write to the file.

Since you have a long term static nameserver, you're good.


Alternatively, if you might change your setup here and there for a
particular host, I find resolvconf to be the perfect middle ground -
and when it's meant to not update resolv.conf, that "NO" setting
looks like it should work a treat for you.

Or if you comission a laptop which might move daily between
locations, fully dynamic DHCP config might be best - just use network
manager and whatever it depends on, and the default "auto everything"
should work assuming DHCP is set up correctly at your site - just let
the auto-stuff do its auto stuff, and the laptop will mostly work.
Uninstall resolvconf since network manager does it all.

If you in this last (laptop) scenario need specific addition of your
static nameserver, on top of the DHCP nameservers (e.g. one
nameserver might resolve some internal names, the others might
recursively resolve internet names) then configuring resolvonf to
know about your static server and add it "on each dhcp
reconfiguration" will work when properly configured too - either
before or after the DHCP assigned servers (i.e. depending on the
precedence you need for the resolves), this could be another ideal
configuration well served by resolvconf.


That which is ideal, always depends on your requirements. I'm still
not clear on your requirements, but perhaps one of the above
scenarios will satisfy.


> On the HP-UX systems, I modified the boot scripts to accomodate all of
> the funky things that happen with DHCP.  I keep master copies of the true
> /etc/resolv.conf and /etc/hosts files (plus one other HP-UX-specific
> network config file), which all get overwritten by the DHCP client,
> and I restore them at just the right time.

That sounds ugly, and way overly hacky.

I suggest thoroughly reading the resolvconf man page to understand
its workings, and making use of that.

If you don't want to do that for any reason, that's also fine - just
remove resolvconf as suggested by someone else in this thread.


> I put the hostname and IP
> in /etc/hosts so that dtlogin can start.  Also, starting with an update
> of ITD's DHCP servers about 5 years ago, the HP-UX DHCP client started
> overwriting the system hostname with the one provided by the DHCP server.
>
> So I had to make more changes to work around that as well, restoring
> the correct hostname at just the right moment.

Surely this would be easier? :
chmod 400 /etc/hostname


I'm a sure fan of static network config (which includes hostname),
but sure sounds like you might be doing things the hard way.


> On the Debian systems, I learned through trial and error that this
> change works:
> 
> --- /etc/dhcp/dhclient.conf.20170410  2017-04-10 10:12:13.271209076 -0400
> +++ /etc/dhcp/dhclient.conf   2017-04-10 10:12:26.291209232 -0400
> @@ -14,8 +14,7 @@
>  
>  send host-name = gethostname();
>  request subnet-mask, broadcast-address, time-offset, routers,
> - domain-name, domain-name-servers, domain-search, host-name,
> - dhcp6.name-servers, dhcp6.domain-search, dhcp6.fqdn, dhcp6.sntp-servers,
> + dhcp6.sntp-servers,
>   netbios-name-servers, netbios-scope, interface-mtu,
>   rfc3442-classless-static-routes, ntp-servers;
> 
> If I simply don't ask for DNS nameserver addresses, the DHCP server
> doesn't provide them, and dhclient doesn't touch resolv.conf, and everyone
> is happy.  The Debian/ISC DHCP client is smart enough to ignore the
> hostname coming from DHCP, so I didn't have to do anything there.
> 
> (The HP-UX DHCP client does not have this level of control, so I have
> to fix things up after the fact.)

Possibly just compile the ISC DHCP client for HP-UX - a binary
compiled by someone else might even be available.


> Home system: the DHCP server is a Belkin plastic router.  The router
> silently forwards DNS queries to the ISP's nameserver, which is tolerable,
> but a bit slow.  I would prefer to run a local dnscache resolver (from
> djbdns).  The DHCP server provides a nameserver entry whether I request
> it or not, and a search domain of "Belkin".
> 
> I am not using djbdns from a Debian package.  It's pure upstream.

openresolv is a new package I haven't seen before

For DNS cache, I've installed dnsmasq many times over the years -
it's small and simple.



> I have configured the router to assign my IP address to my MAC address.
> I just don't want it to touch my resolv.conf file.
> 
> At first I tried doing what I did 

Re: Fonts readability (was: Arial vs. Helvetica.)

[rhkramer]

On Thursday, August 03, 2017 08:23:37 AM Nicolas George wrote:
> Even worse, the anti-aliasing is done wrong:
> it is done without taking gamma correction into account. That means that
> when 50% intensity is wanted, it produces 22% intensity instead:
> black-on-white is too thick, white-on-black is too thin.

Thank you for confirming something I suspected for a long time.

Do you know what program(s) is responsibe for the anti-aliasing, against which 
a bug might be filed (or maybe a bug has already been filed)?

Re: Live recording

[Rodolfo Medina]

Rodolfo Medina  writes:

> Jeremy Nicoll  writes:
>
>> On Thu, 3 Aug 2017, at 13:23, Rodolfo Medina wrote:
>>
>>> 3) mic2 is stereo.
>>
>> Not necessarily.  It might be a mono signal that's been duplicated across
>> two outputs.

Is there, as far as you know, a mono microphone model that we can be sure it
does not do so?  I.e., its sound is heared on only one channel?



>> It's only 2-channel if it can record two separate sounds at the
>> same time.
>>
>> You can get microphones which, in a single unit, record more than one 
>> channel at once.  Their labelling or artwork on them normally makes it 
>> clear that they have L and R (or more than that) inputs.
>>
>> What makes and models are your mics?
>
>
> I can't tell, because they're old and I don't keep their specs.  One of them
> was given to me about 15 years ago, the other I don't even remember how I
> have it.  But I've got another teo: they are, each of them, part of a
> headphone.  They also let hear the recorded sound from both channels.  They
> are:
>
>  http://www.trust.com/it/product/11916-primo-chat-headset
>
> and
>
>  http://www.trust.com/it/product/20685-mauro-headset


Just now I received a reply from Trust assistence and they say the mic (Primo)
is mono.

Anyway, about my example above, can we conclude that mic1 is mono and that the
mic input of my PC is stereo...?

Thanks,

Rodolfo

Re: Virtualisation de très vieux noyaux

[Andre Majorel]

On 2017-08-03 15:09 +0200, Olivier wrote:

> J'ai une connaissance dont l'entreprise fait tourner des
> applications web métiers sur de très vieux noyaux datant de
> 2002, pour situer les choses
> 
> Comme les machines qui hébergent ces applications tombent
> progressivement en panne, ma connaissance souhaite les
> remplacer par des machines virtuelles le temps que le
> remplacement des applications soit décidé, (ce temps là
> pouvant lui aussi se mesurer en années).
> 
> Pour être le plus clair possible, est-il possible de copier un
> disque d'une très vieille machine (disons Debian 2.1 Slink)
> sur un disque d'une VM quelconque (ESXi, KVM, ...) et espérer
> que ça marche pendant encore 1 an ou
> 2 ?
> 
> Avez-vous des retours d'expérience réussie ou non, sur ce type
> d'opération ?

À cette époque, j'installais dans des VMware toutes sortes
d'Unix a priori pas prévus pour la virtualisation. Ici, c'est
pareil sauf qu'il ne s'agit pas d'une nouvelle installation mais
d'une migration.

Le difficultés viendront des périphériques. En particulier le
réseau : il faut que la VM sache émuler une carte ethernet pour
laquelle il existe un driver dans le kernel. De préférence la
même que celle installée dans le PC sur lequel l'OS a été
installé à l'origine.

-- 
André Majorel 
Ever got spam through an address harvested from lists.debian.org ?
Neither have I.

Re: Virtualisation de très vieux noyaux

[Basile Starynkevitch]

On 08/03/2017 03:09 PM, Olivier wrote:

Bonjour,

J'ai une connaissance dont l'entreprise fait tourner des applications 
web métiers sur de très vieux noyaux datant de 2002, pour situer les 
choses


Comme les machines qui hébergent ces applications tombent 
progressivement en panne, ma connaissance souhaite les remplacer par 
des machines virtuelles le temps que le remplacement des applications 
soit décidé, (ce temps là pouvant lui aussi se mesurer en années).


Pour être le plus clair possible, est-il possible de copier un disque 
d'une très vieille machine (disons Debian 2.1 Slink) sur un disque 
d'une VM quelconque (ESXi, KVM, ...) et espérer que ça marche pendant 
encore 1 an ou 2 ?



A mon avis, il serait probablement plus simple de faire tourner la VM 
sur un noyau plus récent (quitte à avoir tout le reste très vieux).


C'est un avis "théorique". Linus a toujours affirmé haut et fort que du 
point de vue applicatif, les noyaux restent compatibles avec des vieux 
exécutables.


(par contre un noyau récent supporte du matériel récent)

Cordialement

--
Basile STARYNKEVITCH   == http://starynkevitch.net/Basile
opinions are mine only - les opinions sont seulement miennes
Bourg La Reine, France

Re: Live recording

[Rodolfo Medina]

Jeremy Nicoll  writes:

> On Thu, 3 Aug 2017, at 13:23, Rodolfo Medina wrote:
>
>> 3) mic2 is stereo.
>
> Not necessarily.  It might be a mono signal that's been duplicated across two
> outputs.  It's only 2-channel if it can record two separate sounds at the
> same time.
>
> You can get microphones which, in a single unit, record more than one 
> channel at once.  Their labelling or artwork on them normally makes it 
> clear that they have L and R (or more than that) inputs.
>
> What makes and models are your mics?


I can't tell, because they're old and I don't keep their specs.  One of them
was given to me about 15 years ago, the other I don't even remember how I have
it.  But I've got another teo: they are, each of them, part of a headphone.
They also let hear the recorded sound from both channels.  They are:

 http://www.trust.com/it/product/11916-primo-chat-headset

and

 http://www.trust.com/it/product/20685-mauro-headset

Rodolfo

Virtualisation de très vieux noyaux

[Olivier]

Bonjour,

J'ai une connaissance dont l'entreprise fait tourner des applications web
métiers sur de très vieux noyaux datant de 2002, pour situer les choses

Comme les machines qui hébergent ces applications tombent progressivement
en panne, ma connaissance souhaite les remplacer par des machines
virtuelles le temps que le remplacement des applications soit décidé, (ce
temps là pouvant lui aussi se mesurer en années).

Pour être le plus clair possible, est-il possible de copier un disque d'une
très vieille machine (disons Debian 2.1 Slink) sur un disque d'une VM
quelconque (ESXi, KVM, ...) et espérer que ça marche pendant encore 1 an ou
2 ?

Avez-vous des retours d'expérience réussie ou non, sur ce type d'opération ?

Par avance, merci.
Slts

Re: Sudden death of bluetooth headphone connectivity

[stuart watt]

On 08/03, Mark Fletcher wrote:

Hello the list!

Suddenly, earlier this week, my bluetooth headphones stopped working
with Stretch.

I update weekly, usually on Sundays, and I am not sure because I wasn't
paying attention whether I had successfully used my headphones before
the last update I did last Sunday.

They definitely were working properly one week before that, when I used
them with the computer just before going off on a week-long business
trip.

Anyway, I use Gnome. When I want to connect the headphones I turn them
on, click on the network-like icon in the top right of the screen, click
on "Not in Use" beside the Bluetooth icon, and then click on "Bluetooth
Settings". (Is there a more efficient, involving fewer clicks, way to
connect, by the way?)

In Bluetooth Settings I can see my headphones as "BeoPlay H8". If I
click on them it brings up the window with the slider switch to connect
them. Normally, what would have happened is that clicking that switch
would have connected the headphones. WHat happens now is the switch goes
very briefly to the ON position, and then goes straight back to the OFF
position. Removing / forgetting the device and re-pairing leads to the
same result -- pairing is successful, but cannot connect.

With journalctl -f running in a terminal while trying to connect I see
the following at the moment of clicking the slider switch to connect the
headphones: a2dp-sink profile connect failed for AA:BB:CC:DD:EE:FF:
Protocol not available


Now, call me a super-sleuth, but I suspect that Protocol Not Available
error could be something to do with the problem ;)

To re-iterate, this was working until recently, and all I've done is
install regular updates using apt update / apt upgrade.

Where should I start looking?

[Full disclosure -- I previously had a different problem on upgrading to
Stretch from Jessie -- see this link for that problem and a link to its
solution --> https://lists.debian.org/debian-user/2017/06/msg01196.html
I have never been particularly happy with the solution I linked to in
that thread, precisely because I was worried it would be damaged by
updates, but it seems to be intact at this point, so I don't think that
is what the problem is here]

Thanks in advance if anyone can help me diagnose the sudden loss of the
a2dp-sink protocol. The headphones work fine with my iPhone, so the
problem seems to be with the computer, with which they were working fine
until a bit over a week ago.

Mark



Pulseaudio? If so what does pavucontrol say

Did Bluez updates come through? maybe pulseaudio-bluetooth module?

regards...


signature.asc
Description: PGP signature

Re: Network config

[Greg Wooledge]

On Thu, Aug 03, 2017 at 11:09:48AM +1000, Zenaan Harkness wrote:
> Perhaps in certain scenarios, yes - the world's full of wierd
> networks :)

Concrete examples?  OK.

Work system: the Information Technology Division (ITD) runs the DHCP
servers, which are basically built around the assumption that you
will be running the standard Windows 7 desktop PC.  The DHCP server
provides nameserver entries if requested, and a hostname whether you
want it or not.

Meanwhile, over in Unixland, my department has its own network
infrastructure which predates the central ITD infrastructure.  We were
the pioneers, not them.  We have our own subdomain (eeg.ccf.org), our
own authoritative DNS servers and caching resolvers.  Our DNS servers
include proper PTR records for our hosts (ITD's do not).

There is a sharing of responsibility between ITD and my department.
They control the allocation of IP addresses, but we are allowed to request
DHCP reservations.  There is a specific protocol for doing so: we put
our new computer on the network, and configure it as a DHCP client.
We get it all set up, with its DHCP-assigned address.  Then we send
an email to the ITD contact address requesting the DHCP reservation.
They put an entry in whatever software they use so the new system *keeps*
its IP address.

If someone moves the machine to a different floor (which will have a
different network address), then the fact that it's configured as a
DHCP client means it will at least *boot* and (hopefully) come up with
a new IP address.  Then they can call me, I can come down to the new
location, get the IP and MAC, update it in DNS, and send the new
DHCP reservation request to ITD.

But the problem is, various Unix DHCP client daemons do *too much*.
All I want them to do is set the IP address, netmask, and gateway.
I *don't* want them to change the system hostname, or the system
resolv.conf (in which I have hand-placed *our* DNS search domain and
*our* DNS resolvers).

On the HP-UX systems, I modified the boot scripts to accomodate all of
the funky things that happen with DHCP.  I keep master copies of the true
/etc/resolv.conf and /etc/hosts files (plus one other HP-UX-specific
network config file), which all get overwritten by the DHCP client,
and I restore them at just the right time.  I put the hostname and IP
in /etc/hosts so that dtlogin can start.  Also, starting with an update
of ITD's DHCP servers about 5 years ago, the HP-UX DHCP client started
overwriting the system hostname with the one provided by the DHCP server.
So I had to make more changes to work around that as well, restoring
the correct hostname at just the right moment.

(Thank goodness for non-parallel boot systems.)

On the Debian systems, I learned through trial and error that this
change works:

--- /etc/dhcp/dhclient.conf.201704102017-04-10 10:12:13.271209076 -0400
+++ /etc/dhcp/dhclient.conf 2017-04-10 10:12:26.291209232 -0400
@@ -14,8 +14,7 @@
 
 send host-name = gethostname();
 request subnet-mask, broadcast-address, time-offset, routers,
-   domain-name, domain-name-servers, domain-search, host-name,
-   dhcp6.name-servers, dhcp6.domain-search, dhcp6.fqdn, dhcp6.sntp-servers,
+   dhcp6.sntp-servers,
netbios-name-servers, netbios-scope, interface-mtu,
rfc3442-classless-static-routes, ntp-servers;

If I simply don't ask for DNS nameserver addresses, the DHCP server
doesn't provide them, and dhclient doesn't touch resolv.conf, and everyone
is happy.  The Debian/ISC DHCP client is smart enough to ignore the
hostname coming from DHCP, so I didn't have to do anything there.

(The HP-UX DHCP client does not have this level of control, so I have
to fix things up after the fact.)


Home system: the DHCP server is a Belkin plastic router.  The router
silently forwards DNS queries to the ISP's nameserver, which is tolerable,
but a bit slow.  I would prefer to run a local dnscache resolver (from
djbdns).  The DHCP server provides a nameserver entry whether I request
it or not, and a search domain of "Belkin".

I am not using djbdns from a Debian package.  It's pure upstream.

I have configured the router to assign my IP address to my MAC address.
I just don't want it to touch my resolv.conf file.

At first I tried doing what I did at work: removing lines from the
dhclient.conf file.  But the Belkin DHCP server doesn't stop sending
the nameserver, and dhclient continues to overwrite resolv.conf.

I ended up simply doing chattr +i.


> If you're wanting a static resolv.conf, are you manually
> running dhclient, or using /etc/network/interfaces ?

I use /etc/network/interfaces with

auto eth0
iface eth0 inet dhcp

> I.e. is this a one-shot scenario, or a "static" config situation?

It's "static" in the sense that the configuration is (supposed to
be) unchanging.  But not in the sense of "iface eth0 inet static".
The IP address, netmask and gateway come from DHCP.  The hostname,
nameservers, and everything else come from their respective 

KDE 5 usando mais memória do que o esperado

[Valentim Carlos]

Bom dia,



  Estou com uma instalação nova do Debian com KDE 5 e esta usando 
mais memória do que o esperado, Em Idle, sem nada aberto esta usando por volta 
de 950 K, coisa que era pra ser entre 400 e 500 K, o que nas minhas outras 3 
maquinas estão usando. Com o Iceweasel aberto e umas 3 páginas esta usando por 
volta de 2,2 GiB, o que nas outras máquinas usa por volta de 1,6 GiB.

   Alguma ideia sobre o que pode ocorrer isso ? pergunto posi tenho 
que rodar Encoder e Render que usa muita memória ...



  Agradeço antecipadamente,



  Valentim Carlos

Sudden death of bluetooth headphone connectivity

[Mark Fletcher]

Hello the list!

Suddenly, earlier this week, my bluetooth headphones stopped working 
with Stretch.

I update weekly, usually on Sundays, and I am not sure because I wasn't 
paying attention whether I had successfully used my headphones before 
the last update I did last Sunday.

They definitely were working properly one week before that, when I used 
them with the computer just before going off on a week-long business 
trip.

Anyway, I use Gnome. When I want to connect the headphones I turn them 
on, click on the network-like icon in the top right of the screen, click 
on "Not in Use" beside the Bluetooth icon, and then click on "Bluetooth 
Settings". (Is there a more efficient, involving fewer clicks, way to 
connect, by the way?)

In Bluetooth Settings I can see my headphones as "BeoPlay H8". If I 
click on them it brings up the window with the slider switch to connect 
them. Normally, what would have happened is that clicking that switch 
would have connected the headphones. WHat happens now is the switch goes 
very briefly to the ON position, and then goes straight back to the OFF 
position. Removing / forgetting the device and re-pairing leads to the 
same result -- pairing is successful, but cannot connect.

With journalctl -f running in a terminal while trying to connect I see 
the following at the moment of clicking the slider switch to connect the 
headphones: a2dp-sink profile connect failed for AA:BB:CC:DD:EE:FF: 
Protocol not available


Now, call me a super-sleuth, but I suspect that Protocol Not Available 
error could be something to do with the problem ;)

To re-iterate, this was working until recently, and all I've done is 
install regular updates using apt update / apt upgrade.

Where should I start looking?

[Full disclosure -- I previously had a different problem on upgrading to 
Stretch from Jessie -- see this link for that problem and a link to its 
solution --> https://lists.debian.org/debian-user/2017/06/msg01196.html 
I have never been particularly happy with the solution I linked to in 
that thread, precisely because I was worried it would be damaged by 
updates, but it seems to be intact at this point, so I don't think that 
is what the problem is here]

Thanks in advance if anyone can help me diagnose the sudden loss of the 
a2dp-sink protocol. The headphones work fine with my iPhone, so the 
problem seems to be with the computer, with which they were working fine 
until a bit over a week ago.

Mark

Re: Live recording

[Jeremy Nicoll]

On Thu, 3 Aug 2017, at 13:23, Rodolfo Medina wrote:

> 3) mic2 is stereo.

Not necessarily.  It might be a mono signal that's been duplicated
across 
two outputs.  It's only 2-channel if it can record two separate sounds
at
the same time. 

You can get microphones which, in a single unit, record more than one 
channel at once.  Their labelling or artwork on them normally makes it 
clear that they have L and R (or more than that) inputs.

What makes and models are your mics?

-- 
Jeremy Nicoll - my opinions are my own.

Fonts readability (was: Arial vs. Helvetica.)

[Nicolas George]

Le sextidi 16 thermidor, an CCXXV, Martin Read a écrit :
> On a computer screen, I tend to find that sans or quasi-sans (e.g. fonts
> where 'I' and 'l' have serifs but other letters mostly don't) fonts are more
> comfortable to read (and, in particular, hold up better at small point
> sizes).

At small point sizes, the bitmap fonts still beat the vectorial fonts in
terms of readability.

For my XTerms, I use courR12.pcf.gz from xfonts-75dpi, aka
-adobe-courier-medium-r-normal--12-120-75-75-m-70-iso10646-1; its glyphs
are 7×13 pixels. I would not change it for a vectorial font. I sometimes
also use 7x13.pcf.gz from xfonts-base, aka
-misc-fixed-medium-r-normal--13-120-75-75-c-70-iso10646-1 because it is
more complete.

Notice that -adobe-courier-* is a serif font while -misc-fixed- is a
sans-serif font. But I find -adobe-courier-* much more readable,
especially for long text (and when I say long, I mean MBOTF-long).

Another point where the bitmap fonts beat the vectorial fonts at tiny
sizes: you usually want your vectorial fonts anti-aliased, but at tiny
sizes it hurts readability. Even worse, the anti-aliasing is done wrong:
it is done without taking gamma correction into account. That means that
when 50% intensity is wanted, it produces 22% intensity instead:
black-on-white is too thick, white-on-black is too thin.

Regards,

-- 
  Nicolas George


signature.asc
Description: Digital signature

Re: Live recording

[Rodolfo Medina]

deloptes  writes:

> Rodolfo Medina wrote:
>
>> Now I was wondering about the stereo o non-stereo character of such a home
>> made recording...  I tried to use two microphones together, plugging them
>> together into the PC with a small common connection doubber.  Can we say
>> the result is stereo...?  I would doubt...  and how to have - if possible
>> - a stereo effect with the above basic recording instruments?
>
> [...]
>
> The mic is mono, so the input if I am not wrong is mono as well.
>
> [...]
>
> The mic has also different Ohms or perhaps better to say impedance, so I am
> not sure if it would work, but you could try plugging those mics to the
> line in, not the mic.
>
> So line in, is the stereo one. I think if you plugin mic(s) without preamp,
> because of the high Ohms, it would be very silent, but still worth a try.
>
> You need a stereo jack to L-R like this one
> https://www.thomann.de/at/pro_snake_78219_yadapterkabel.htm
>
> This way you'll also know if it is true stereo, because if you unplug the
> one mic, it would record only on the other channel. Of course you can see
> this in audacity or even with arecord


With my two microphones, say mic1 and mic2, I did a little experiment: I
recorded my voice with mic1, using sox, then listened to the so created .wav
file and the sound was heard only on the left channel.  Instead, doing the same
with mic2, the sound was heard on both channels.  How should we conclude?  It
seems to me that we should conclude that:

1) the mic input on my PC is stereo.  In fact, it is a laptop, nay a netbook,
   doesn't have a line in and it is reasonable that its mic input is also a
   line in;

2) mic1 is mono;

3) mic2 is stereo.

Do you agree?  But then, if it is so, my rough home made live piano record,
done using mic1 and mic2 plugged together in a one-male--two-female splitter,
*is* actually stereo...?

Thanks,

Rodolfo

Re: Arquitetura i386 para amd64

[Valentim Carlos]

Amigo, Eu "Quase" consegui ... rsrs ...



Bom, como eu precisava do sistema em 64 e ia ter que reinstalar novamente, eu tentei o cross-grade. 
Bom o kernel eu já tinha mudado para amd64, ai fiz os procedimentos e tal ... consegui mudar a 
arquitetura "não estrangeira" para amd64 e a arquitetura "estrangeira" para 
i386, ai devo devo ter errado em algum ponto, pois não conseguia instalar alguns pacotes 64 ...

De qualquer forma, fui muito, mais muito útil, agora reinstalei o sistema com 
Debian Sid pelo mini.iso - expert install ...



Obrigado novamente ...



Valentim Carlos


Em 01 de ago de 2017 às 14:26, Henrique de Moraes Holschuh  
escreveu:


Olha, trocar um Debian de i386 para amd64 é possível via "Multiarch".
Chamamos isso de "cross-grade". Mas é uma dor de cabeça que não é
recomendada nem para quem é usuário avançado.

Se for simples reinstalar, reinstale. Se não for, e quiser tentar um
cross-grade, *FAÇA BACKUP FULL INCLUSIVE DO SISTEMA* e boa sorte (porque
vai precisar):

https://wiki.debian.org/CrossGrading
http://collab.debian.net/portal/planet-debian/jose-m.-calhariz-crossgrading-a-complex-desktop-and-debian-developer-machine-running-debian-9

--
Henrique Holschuh

Re: plusieurs ports d'écoute pour un serveur SSH

[Thierry Bugier Pineau]

Le jeudi 03 août 2017 à 11:12 +0200, Samy Mezani a écrit :
> Le 03/08/2017 à 10:56, daniel huhardeaux a écrit :
> > Le 03/08/2017 à 10:50, Samy Mezani a écrit :
> > > 
> > > 
> > > Le 03/08/2017 à 10:26, Yatajb a écrit :
> > > > Un firewall ?
> > > 
> > > Je n'ai pas bidouillé le pare-feu, donc par défaut tous les ports
> > > sont 
> > > accessibles non ?
> > > 
> > > Voici un extrait de la config du pare-feu sur le serveur :
> > > # iptables -L
> > > Chain INPUT (policy DROP)
> > 
> > Euh ... ports accessibles avec cette policy ;), m'étonnerai
> > beaucoup !
> > 
> > > [...]
> > > ACCEPTtcp  --  anywhereanywhere tcp dpt:ssh
> > > [...]
> > 
> > ssh est le service soit le port 22. Il faut rajouter ton port.
> 
> OK, merci, j'ignorais qu'il fallait explicitement ouvrir ce port.
> Pas 
> trop l'habitude de manipuler mon pare-feu…
> 
> J'ai résolu mon problème avec :
> # iptables -A INPUT -p tcp --dport  -j ACCEPT
> 
> Merci à vous 2 !
> 
> Samy
> 

Bonjour

Pour info il existe ufw (Uncomplicated firewall) sous Debian, avec une
petite interface graphique pour Gnome. J'ai pu m'en servir brièvement
quand je connectais mon pc poratble sur un réseau bourré de machines
windows. De mémoire c'est une couche d'abstraction au dessus d'iptable,
et créé par Canonical / Ubuntu. J'ai trouvé que le rapport complexité /
fonctionnalité était pas mal du tout. A essayer :)

Re: plusieurs ports d'écoute pour un serveur SSH

[Yatajb]

You're welcome :-) 


Jean-Bernard YATA via mobile

Groupe ACE SI
Consultant Sécurité des Systèmes d'Information
Lead Auditor ISO 27001
Pentester | Hackeur Éthique 
Enseignant Universitaire en Sécurité Informatique

jby...@ace-si.com | 06.24.41.18.27 | www.ace-si.com

> Le 3 août 2017 à 11:12, Samy Mezani  a écrit :
> 
>> Le 03/08/2017 à 10:56, daniel huhardeaux a écrit :
>>> Le 03/08/2017 à 10:50, Samy Mezani a écrit :
>>> 
>>> 
 Le 03/08/2017 à 10:26, Yatajb a écrit :
 Un firewall ?
>>> 
>>> Je n'ai pas bidouillé le pare-feu, donc par défaut tous les ports sont 
>>> accessibles non ?
>>> 
>>> Voici un extrait de la config du pare-feu sur le serveur :
>>> # iptables -L
>>> Chain INPUT (policy DROP)
>> Euh ... ports accessibles avec cette policy ;), m'étonnerai beaucoup !
>>> [...]
>>> ACCEPTtcp  --  anywhereanywhere tcp dpt:ssh
>>> [...]
>> ssh est le service soit le port 22. Il faut rajouter ton port.
> 
> OK, merci, j'ignorais qu'il fallait explicitement ouvrir ce port. Pas trop 
> l'habitude de manipuler mon pare-feu…
> 
> J'ai résolu mon problème avec :
> # iptables -A INPUT -p tcp --dport  -j ACCEPT
> 
> Merci à vous 2 !
> 
> Samy
> 

Re: plusieurs ports d'écoute pour un serveur SSH

[Samy Mezani]

Le 03/08/2017 à 10:56, daniel huhardeaux a écrit :

Le 03/08/2017 à 10:50, Samy Mezani a écrit :



Le 03/08/2017 à 10:26, Yatajb a écrit :

Un firewall ?


Je n'ai pas bidouillé le pare-feu, donc par défaut tous les ports sont 
accessibles non ?


Voici un extrait de la config du pare-feu sur le serveur :
# iptables -L
Chain INPUT (policy DROP)


Euh ... ports accessibles avec cette policy ;), m'étonnerai beaucoup !


[...]
ACCEPTtcp  --  anywhereanywhere tcp dpt:ssh
[...]


ssh est le service soit le port 22. Il faut rajouter ton port.


OK, merci, j'ignorais qu'il fallait explicitement ouvrir ce port. Pas 
trop l'habitude de manipuler mon pare-feu…


J'ai résolu mon problème avec :
# iptables -A INPUT -p tcp --dport  -j ACCEPT

Merci à vous 2 !

Samy

Re: plusieurs ports d'écoute pour un serveur SSH

[Yatajb]

Oui tu as spécifié :ssh 
Ce qui signifie explicitement le port 22 uniquement.
Rajoute la même avec : 


Jean-Bernard YATA via mobile

Groupe ACE SI
Consultant Sécurité des Systèmes d'Information
Lead Auditor ISO 27001
Pentester | Hackeur Éthique 
Enseignant Universitaire en Sécurité Informatique

jby...@ace-si.com | 06.24.41.18.27 | www.ace-si.com

> Le 3 août 2017 à 10:50, Samy Mezani  a écrit :
> 
> 
> 
>> Le 03/08/2017 à 10:26, Yatajb a écrit :
>> Un firewall ?
> 
> Je n'ai pas bidouillé le pare-feu, donc par défaut tous les ports sont 
> accessibles non ?
> 
> Voici un extrait de la config du pare-feu sur le serveur :
> # iptables -L
> Chain INPUT (policy DROP)
> [...]
> ACCEPTtcp  --  anywhereanywhere tcp dpt:ssh
> [...]
> 
> Samy
> 

Re: plusieurs ports d'écoute pour un serveur SSH

[daniel huhardeaux]

Le 03/08/2017 à 10:50, Samy Mezani a écrit :



Le 03/08/2017 à 10:26, Yatajb a écrit :

Un firewall ?


Je n'ai pas bidouillé le pare-feu, donc par défaut tous les ports sont 
accessibles non ?


Voici un extrait de la config du pare-feu sur le serveur :
# iptables -L
Chain INPUT (policy DROP)


Euh ... ports accessibles avec cette policy ;), m'étonnerai beaucoup !


[...]
ACCEPTtcp  --  anywhereanywhere tcp dpt:ssh
[...]


ssh est le service soit le port 22. Il faut rajouter ton port.

--
Daniel

Re: plusieurs ports d'écoute pour un serveur SSH

[Samy Mezani]

Le 03/08/2017 à 10:26, Yatajb a écrit :

Un firewall ?


Je n'ai pas bidouillé le pare-feu, donc par défaut tous les ports sont 
accessibles non ?


Voici un extrait de la config du pare-feu sur le serveur :
# iptables -L
Chain INPUT (policy DROP)
[...]
ACCEPTtcp  --  anywhereanywhere tcp dpt:ssh
[...]

Samy

Re: plusieurs ports d'écoute pour un serveur SSH

[Yatajb]

Un firewall ? 


Jean-Bernard YATA via mobile

Groupe ACE SI
Consultant Sécurité des Systèmes d'Information
Lead Auditor ISO 27001
Pentester | Hackeur Éthique 
Enseignant Universitaire en Sécurité Informatique

jby...@ace-si.com | 06.24.41.18.27 | www.ace-si.com

> Le 3 août 2017 à 10:18, Samy Mezani  a écrit :
> 
> Bonjour,
> 
> J'ai un serveur ssh qui fonctionne habituellement sur le port classique 22.
> Je souhaite dans l'immédiat ajouter un autre port d'écoute, par exemple le 
> .
> 
> J'ajoute la ligne 'Port ' sous la ligne 'Port 22' dans le fichier 
> /etc/sshs/sshd_config.
> Je redémarre le serveur ssh, je vérifie les ports sur lequel il écoute avec 
> netstat et j'ai bien :
> tcp 0  0 0.0.0.0: 0.0.0.0:*LISTEN   10710/sshd
> 
> Sur le serveur lui même, la commande 'ssh -p  user@localhost' fonctionne.
> 
> Mais depuis un poste client sur le même réseau local, la commande 'ssh -p 
>  user@192.168.X.X' n'aboutit pas et j'obtiens l'erreur : "ssh: connect to 
> host 192.168.X.X port : Connection timed out"
> 
> Par contre la commande 'ssh -p 22 user@192.168.X.X' fonctionne.
> 
> Je ne comprends pas. Qu'est-ce que je fais mal ?
> 
> Samy
> 

Re: Arial vs. Helvetica.

[Martin Read]

On 02/08/17 16:34, Joe wrote:

Incidentally, serifs were invented to make blocks of text easier to
read, so Times or similar would be a better choice for paragraphs, with
a sans-serif font more suited to bold headings. Newspaper sites (not
surprisingly including The Times) use serif fonts.


On a computer screen, I tend to find that sans or quasi-sans (e.g. fonts 
where 'I' and 'l' have serifs but other letters mostly don't) fonts are 
more comfortable to read (and, in particular, hold up better at small 
point sizes).

plusieurs ports d'écoute pour un serveur SSH

[Samy Mezani]

Bonjour,

J'ai un serveur ssh qui fonctionne habituellement sur le port classique 22.
Je souhaite dans l'immédiat ajouter un autre port d'écoute, par exemple 
le .


J'ajoute la ligne 'Port ' sous la ligne 'Port 22' dans le fichier 
/etc/sshs/sshd_config.
Je redémarre le serveur ssh, je vérifie les ports sur lequel il écoute 
avec netstat et j'ai bien :

tcp 0  0 0.0.0.0: 0.0.0.0:*LISTEN   10710/sshd

Sur le serveur lui même, la commande 'ssh -p  user@localhost' 
fonctionne.


Mais depuis un poste client sur le même réseau local, la commande 'ssh 
-p  user@192.168.X.X' n'aboutit pas et j'obtiens l'erreur : "ssh: 
connect to host 192.168.X.X port : Connection timed out"


Par contre la commande 'ssh -p 22 user@192.168.X.X' fonctionne.

Je ne comprends pas. Qu'est-ce que je fais mal ?

Samy

Re: Re: apt or apt-get equivalent of 'aptitude --show-why upgrade'

[Christoph Groth]

Thanks for the prompt reply!

The Wanderer wrote:


On 2017-08-02 at 11:44, Christoph Groth wrote:



> I'm running Debian testing and would like to upgrade from
> "oldtesting" (jessie) to current testing.  I noticed that 'apt
> upgrade' as well as 'apt-get upgrade' want to install the 
> package

> 'pulsaudio' which I've been avoiding successfully so far.

That's a little weird. According to the apt-get man page, 
'upgrade' will

never install a package which is not already installed.


Sorry, I was a bit imprecise.  The three tools (apt-get, aptitude 
& apt) behave in slightly different ways.  I should have simply 
asked whether there is any way (other than aptitude's option) to 
know why something is about to be installed either by "upgrade" or 
"dist-upgrade"


The 'upgrade' command for apt behaves differently; that one will 
install
new packages if needed, but will never remove an existing 
package.


Are you certain that both commands produce the same 
install-new-package

result?

> I could of course uninstall pulseaudio after the upgrade, but 
> I

> wonder whether a more elegant solution does not exist.

Try 'apt-get upgrade pulseaudio-'; the trailing hyphen should 
tell
apt-get to remove the package, which in this case means not 
installing
it. I more usually use this sort of thing with dist-upgrade, but 
AFAIK

it should work for upgrade as well.


Thanks, this works (with a small modification).  If I execute 
'apt-get full-upgrade pulseaudio-' APT suddenly wants to install 
pulsaudio:i386 along with dozens of other i386 packages.  (I'm 
aware of multiarch and use it, but didn't know that packages from 
one arch can substitute others.)  Executing 'apt-get full-upgrade 
pulseaudio- pulseaudio:i386-' does the trick.

Re: determining which apps have entries in the applications menu in xfce

[Joe]

On Wed, 2 Aug 2017 21:54:48 -0700
Dan Hitt  wrote:


> 
> I did have trouble with Settings > Main Menu  my version of xfce
> doesn't seem to have that.
> The Settings submeu jumps from Keyboard to MIME Type Editor, and i
> don't see an xfce icon i can click on.  (I also tried the Settings
> Manager, but similarly couldn't find anything.)  My version of xfce is
> version 4.12.
> 
> It sounds like it would be very useful to get to Settings > Main
> Menu, though.
> 

That would be alacarte, a Gnome program but without many dependencies.

-- 
Joe